Lower Facial Reanimation Techniques Following Cancer Resection and Free Flap Reconstruction

Transcription

1 The Laryngoscope VC 2016 The American Laryngological, Rhinological and Otological Society, Inc. Lower Facial Reanimation Techniques Following Cancer Resection and Free Flap Reconstruction Alexandra E. Kejner, MD; Eben L. Rosenthal, MD Objectives/Hypothesis: Evaluate outcomes of the standard static sling and orthodromic temporalis tendon transfer reanimation for facial nerve paralysis. Study Design: Retrospective case series at a tertiary care hospital of head and neck cancer patients with facial nerve palsy secondary to malignancy or resection. Methods: From 2004 to 2014, patients undergoing resection of malignancy that involved facial nerve palsy requiring facial reanimation were identified. All procedures were performed by the senior author (E.L.R.). Demographics, methods, revision rates, combination with other procedures, and complications were evaluated. Results: A total of 77 patients underwent 92 procedures, with two patients requiring more than one revision, for a total of 20 revisions. Average time to revision was 9 months. Age, sex, race, side of repair, paralysis prior to procedure, sling type or method, timing of procedure, and radiation therapy were not significantly different between those requiring revision and those who did not. There was no difference in complications between patients who received radiation and those who did not (P 5.5), nor between static versus orthodromic temporalis muscle transfer (P 5.5). Complication rate was low at 5.4%. Conclusions: Sling procedures can be successfully performed in patients with facial nerve palsy secondary to cancer resection with radiation therapy, with a low revision rate and few complications. Key Words: Facial reanimation, radiation therapy, facial sling, orthodromic temporalis sling. Level of Evidence: 4 Laryngoscope, 126: , 2016 INTRODUCTION Facial nerve paralysis can be a devastating sequelum after cancer resection and can severely impact patient quality of life due to cosmesis, effects on social interaction, and function, in particular eye closure and oral competence. 1 This can cause significant psychological distress in addition to its functional implications. 2 Nerve damage can be secondary to the malignancy itself or to the oncologic resection. Regardless of the underlying pathology, patients have been shown to benefit from reanimation. 3 Reanimation procedures include anything from static procedures to the use of free muscle transfer with cross-facial nerve grafting. 4 Multiple methods of reanimation have been proposed including combinations of resorbable and nonresorbable materials, synthetics, allografts of tendon or fascia lata, free tissue transfers, and cross-facial nerve grafting. 5 7 Choice of facial suspension type is affected by multiple issues including loss From the Department of Surgery, Division of Otolaryngology Head and Neck Surgery, and the Department of Medicine, Division of Immunology and Rheumatology, University of Alabama at Birmingham, Birmingham, Alabama, U.S.A. Editor s Note: This Manuscript was accepted for publication December 7, The authors have no funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Alexandra E. Kejner, MD, Yale Section of Otolaryngology, 800 Howard Ave 4th Floor, New Haven, CT or Eben Rosenthal, Professor of Otolaryngology and Radiology, Ann & John Doerr Medical Director, Stanford Cancer Center, 875 Blake Wilbur Drive, Stanford, CA , elr@stanford.edu DOI: /lary of cranial nerve V, resection of the temporalis muscle or mandible, need for external carotid sacrifice, very large volume of soft tissue extirpation, patient factors, and radiation therapy. 8 Patients undergoing resection for cancer, the majority of whom undergo free flap reconstruction and radiation therapy, are at an increased risk of wound-healing issues due to hypoxia and fibrosis following radiotherapy. Radiation affects wound healing in a number of different ways including skin atrophy, soft tissue fibrosis, desquamation, epithelial ulceration, fistula formation, vessel rupture, and impaired peri- and postoperative wound healing. 7 In these patients, free muscle transfer for reanimation may not be a viable option due to age, comorbidities, and the need for a larger tissue volume than that afforded by a gracilis flap to fill a defect after oncologic resection in preparation for radiation. Additionally, in the setting of advanced-stage malignancy and complete obliteration of the muscular end organ and/or facial nerve trunk, the options for complex nerve reconstruction are more limited. Studies examining techniques in this group have not been widely performed and are often limited by sample size. Standard static sling and orthodromic temporalis muscle transfer (OTM) allow for a one-step procedure that can be performed either on an outpatient basis or concomitantly with extirpation, with a low complication rate and which can be revised as necessary. 5,7 Comparison of these techniques has been difficult to facilitate, as surgical planning is often patient, disease, and surgeon specific. 8 In this study, we evaluated one of

2 Fig. 1. Postoperative photo of the patient 6 months after an orthodromic temporalis sling procedure with his face at rest. [Color figure can be viewed in the online issue, which is available at the largest series of patients with facial nerve palsy after radical parotidectomy and radiotherapy for malignancy to determine the need for and timing of repeat reanimation procedures as well as to document complication rates. MATERIALS AND METHODS Institutional review board approval was obtained from the University of Alabama at Birmingham s board. From 2004 to 2013, patients undergoing resection for carcinoma that involved facial nerve paralysis either preoperatively or that was sacrificed as part of surgical resection were considered for facial reanimation procedures. All procedures were performed by the senior author (E.L.R.). Static Sling The technique consists of an incision made approximately 3 cm in length on the anterior aspect of the free flap or in the melolabial crease. A 1-cm elliptical portion can be removed if debulking of a previously placed flap is necessary. The dissection is then carried anteriorly. The orbicularis oris muscle is identified, and two sutures (most commonly in this series 2-O PDS) are Fig. 3. Postoperative photo of the patient 6 months after a static sling procedure with his face at rest. [Color figure can be viewed in the online issue, which is available at placed through the muscle and secured to the temporalis fascia and then secured to the periosteum of the arch in the malar eminence. Alternatively, a 2-mm burr can be used to drill a hole into the zygoma to suspend Prolene or PDS sutures if the periosteum does not have sufficient integrity to hold the sutures. These are then over-tightened to produce some showing of the ipsilateral canine. The incision is then irrigated, hemostasis is obtained, and the incision is then closed using 3-O Vicryl suture and 4-O plain gut (Figs. 1 and 2). The same technique can be performed simultaneously with free flap, which, if performed in patients with skin deficit, is facilitated as the zygoma and overlying periosteum is easily exposed. Orthodromic Temporalis Muscle Transfer An incision is made in the melolabial crease and dissection carried down to the coronoid process. Once the coronoid process is identified, a clamp is placed on it and a saw is used to divide the superior aspect, thus leaving the coronoid attached to the temporalis muscle. A hole is then drilled through the coronoid process, and 2-O Prolene sutures are secured through the bone. Once this is performed, the 2-O Prolene sutures are secured to the modiolus and to the fascia of the orbicularis oris muscle. These are then over-tightened. A cuff of skin can be removed from the incision to minimize the amount of redundant tissue, and the incision is closed using resorbable suture. This technique does allow for the possibility of dynamic movement with appropriate physical therapy and usually results in about 2 to 4 mm of excursion (Figs. 3 and 4). This technique was described by Boahene et al. 9 When being performed simultaneously with free flap reconstruction, the same melolabial incision can be used, as this gives the benefit of additional lift by incising/excising and tightening the skin of the melolabial crease. Alternatively, if there is already a parotidectomy incision or substantial skin loss over the cheek, the coronoid process may be approached from the existing incision. Skin flaps are elevated toward the modiolus (inferiorly and anteriorly) from this direction. Blunt dissection is carried out until the coronoid process is exposed with its temporalis insertion. The rest of the procedure then continues as described above. Fig. 2. Postoperative photo of the patient 6 months after an orthodromic temporalis sling procedure showing an active smile with about 3 mm of excursion. [Color figure can be viewed in the online issue, which is available at Facial Nerve Grafting When possible, nerve grafting is undertaken using either the medial antebrachial cutaneous nerve or the nerve to vastus lateralis, depending on the donor site. Nerve grafting is only possible when there is sufficient proximal nerve stump as well as intact distal branches. As in the case of patients with 1991

3 TABLE I. Timing of Radiation Therapy. Timing of XRT No. Prior 17 Post 50 Both 3 None 23 XRT 5 radiation therapy. Fig. 4. Postoperative photo of the patient 6 months after a static sling procedure. The patient was asked to smile. [Color figure can be viewed in the online issue, which is available at significant skin and muscle resection due to extirpation of primary tumor, nerve grafting is not a viable option as the end muscle is no longer present. The nerve is harvested separately from the flap itself. The donor nerve is then sutured to the proximal stump and ideally to at least one upper division branch and one lower division branch if available. The use of neuromonitoring can aid in choosing which distal branches are appropriate for grafting. Outcome Analysis Patient demographic data, defect, timing of reconstruction, reconstruction method, length of stay, pre- or postoperative radiation therapy, and need for revision were evaluated and compared. Patients requiring revision were then further evaluated by reconstruction technique and patient characteristics. Need for revision was decided in combination with the surgeon and the patient based on patient perception of facial symmetry and is used as a surrogate for failure of the facial reanimation technique. Due to the retrospective nature of this study, clinical exam and documentation were primarily subjective and based on patient and surgeon assessment of worsened facial asymmetry and dysfunction including cheek biting when eating. As ultimately this is primarily a cosmetic issue that is being addressed, we are using the revision rate as a surrogate objective measure, as it indicates that the patient and surgeon agree that the lower facial suspension is no longer acceptable to the patient. Although excursion might be interesting to note, the overall change may still not reflect patient satisfaction with the suspension, and the patient may still request further revision. Complications included any type of intervention (besides revision alone) including the need for antibiotics, admission with intravenous antibiotics, surgical debridement, or local wound care in the office. Statistical Analyses Descriptive variables were summarized by number (%) for categorical variables. A one-way analysis of variance was used to analyze relationships between categorical factors and continuous responses. A Student t test was used to compare differences in means between groups. A contingency analysis was used to analyze relationships between categorical factors and responses. A P value of <.05 was considered statistically significant. Statistical analysis was performed using Jmp 11 software (SAS Institute Inc., Cary, NC). RESULTS A total of 77 patients underwent a total of 92 lower facial reanimation procedures and were included in the study. Follow-up ranged from 0 to 56 months, with an average of 10.5 months. Average age was 68 years old (range, years old). There were 17 patients who underwent OTM and 60 who underwent the static sling procedure. Preexisting facial paralysis was present in 25 (27%) patients. There were 36 (%) patients who underwent lower facial reanimation at the same time as their resection and flap reconstruction. Their average length of stay was 5.8 days. The remaining patients had their procedures performed as outpatients. Of the 77 patients, 59 (77%) underwent radiation therapy (Table I). All but four patients underwent free flap reconstruction. Most patients underwent multiple adjunctive procedures including gold weight, canthoplasty, and or brow lift. Facial nerve grafting was performed in 23 patients. As there was no consistent documentation of the House-Brackmann scale, excursion was not evaluated in this study. For the patients who did undergo facial nerve grafting, five (22%) required revision surgery, and one of these patients required a second revision surgery. There were no significant differences between the OTM and static sling groups with regard to need for revision or complications (P 5.64). However, there was a significant difference with regard to simultaneous free flap reconstruction between those who underwent OTM versus static sling procedures. All but two of the OTM TABLE II. Demographic Data Comparing Patients Requiring Revision Versus Those Who Did Not (N 5 92). Revision, n 5 20 No Revision, n 5 72 P Value Average age, yr M/F 15/5 65/7.09 Right-side surgery 7 (35%) 38 (53%).16 Static sling 17 (85%) 58 (77%).64 Paralysis prior 4 (20%) 21 (29%).4 XRT1 14 (70%) 56 (78%).48 Simultaneous FF 11 (55%) 30 (42%).29 Four patients did not undergo FF reconstruction. F 5 female, FF 5 free flap; M 5 male, XRT 5 radiation therapy. 1992

4 TABLE III. Characteristics of Patients Requiring Revision. Age, yr Sex Side Time to Revision, mo Paralysis Prior/Post Sling Type Flap Defect Subsite XRT Same Time as Flap 65 M L 17 Post Dynamic No Parotid Post Y 56 M R 13 Post Dynamic Rectus LTB Post Y 64 M R 15 Prior Dynamic Rectus Parotid Post Y 70 F L 9.5 Post Static ALT/rectus LTB No N 74 M L 0.83 Post Static ALT Parotid Prior Y 50 M L 5.3 Prior Static Rectus LTB Post Y 64 M R 2.8 Post Static RFFF Parotid No N 36 F R 5.5 Prior Static Rectus LTB Post Y 76 F L 13 Post Static ALT LTB Post Y 76 F L 13 Post Static ALT LTB No N 67 M L 3.5 Post Static ALT LTB Prior Y 79 M R 1.3 Post Static ALT Parotid No N 71 M L 5.3 Post Static RFFF Parotid Post Y 71 M L 9.3 Post Static RFFF Parotid Post Y 80 M L 2.8 Post Static RFFF Parotid No N 75 M R 1 Post Static RFFF Parotid Post Y 64 M L 26 Prior Static ALT Parotid Post Y 50 F R 2.6 Post Static ALT Parotid Prior Y 61 M L 24 Post Static RFFF Cheek No N 79 M L 11 Post Static ALT Parotid Post Y ALT 5 anterolateral thigh free flap; F 5 female; L 5 left; LTB 5 lateral temporal bone; M 5 male; N 5 no; R 5 right; RFFF 5 radial forearm free flap; XRT 5 radiation therapy; Y 5 yes. patients underwent their facial reanimation at a later date than their original extirpative surgery (P ). Comparing patients who required revision and those who did not, there were no significant differences with regard to age, gender, surgery side, method of suspension, preoperative paralysis, simultaneous free flap or radiation therapy (Table II), nor was there any difference regarding suture type, suture size, number of sutures, or facial nerve grafting. Characteristics of patients undergoing revision are demonstrated in Table III. All revisions were performed with static technique. In the six OTM patients with measurable movement, the average excursion was between 2 to 4 mm. None of these patients had facial nerve grafting, and none of these six patients required a revision. Overall success rate was 77% and was comparable to other similar patient populations (Table IV). TABLE IV. Revision Rates Compared to the Literature. Revision 6 Complication No. XRT1 Method P Value UAB, n (23%) 70 (75%) OTM/static Griffin et al., n (14%) 7 (50%) OTM.73 Leckenby et al., n (8%) 0 (0%) Static.037 Revenaugh et al., n (0%) 4 (80%) OTM.58 Skourtis et al., n (35%) 17 (100%) Static.36 OTM 5 orthodromic temporalis muscle transfer; UAB 5 University of Alabama at Birmingham; XRT 5 radiation therapy. DISCUSSION Patients with advanced-stage disease requiring radical parotidectomy represent a unique challenge due to age, unclear long-term survival, and perioperative radiation therapy with its subsequent effects on healing. We evaluated the success rate of lower facial reanimation procedures utilizing OTM and the static sling technique. A total of 92 procedures were performed for facial reanimation after cancer. Most patients had the procedure performed as an outpatient. In our patients, the only significant difference between the OTM sling and the static sling was concomitant free flap reconstruction of the defect (P ). This was typically performed due to its ease of performance at the same time as surgery as well as the character of the defect, in cases where temporalis muscle was included as part of the specimen or sacrifice of the mandibular branch of the trigeminal nerve. In isolated cases, patients who had undergone previous irradiation also had compromised muscle tone of the temporalis muscle, compelling the surgeon to choose the static sling technique. Additionally, as margin status was often pending at the time of surgery, this may have been deemed to be the more conservative option to avoid issues with the need for re-resection. Wound complications were low, with only five patients (5.4%) having infection, and of these only three required surgical debridement. This is slightly lower compared to other series in similar patients, which have shown a complication rate as high as 17%. 7 Compared to patients undergoing lower facial reanimation procedures 1993

5 for causes other than malignancy (acoustic neuroma, benign parotid neoplasm), there is an increased revision rate and complication rate, which is likely inherent to the patient population and tissue quality. 3 The need for revision was not predicted by patient demographics, sling type, radiation therapy, timing of reanimation surgery, or defect subsite. Revision rates were approximately 22% for all procedures and were not significantly different between radiated and nonradiated patients nor between sling types. This is comparable to findings in similar patient series (range, 5 51 patients) with a revision rate ranging 0% to 35%. 3,7,8,10 Due to the retrospective nature of this study, objective data regarding the decision-making process for revision were not available and were based on patient satisfaction and surgeon assessment of facial flaccidity as well as function, including issues with cheek biting when eating. It would be ideal to address this in the future with prospective study. Gracilis flap was not utilized in this patient population because often a large volume deficit was present, requiring a larger flap. A flaw of this study is that the facial nerve grafting outcomes were not examined more in-depth, as this was a retrospective study, and the postoperative House-Brackmann data were not recorded in every chart. There was no difference in the revision rates for these 23 patients compared to those who did not have grafting performed. However, the number of patients undergoing grafting was small, and future endeavors should include a subset analysis of these patients. Reanimation procedures afford the cancer patient not only functional rehabilitation but also cosmetic improvement for the paralyzed face. Static and dynamic procedures can be safely performed either at the time of initial reconstruction with microvascular free tissue transfer or as an adjunctive procedure later as an outpatient. Revision rates are low even in this group of patients who often require radiation therapy. Choosing which method to use is often dependent on several patient characteristics, as outlined by Skourtis et al., including whether mandible, temporalis, or external carotid have been resected, or if a very large soft tissue defect is present. Static suspension has been shown to be an acceptable reconstructive option, with a 90% to 95% success rate in most cases. 8 Orthodromic temporalis tendon suspension, which does allow for the possibility of dynamic movement with extensive physical therapy, has also been successfully performed in irradiated patients, although wound complications are higher in the radiated population. 7 In our group, there was no difference with regard to revision between the two techniques. They were both acceptable reanimation techniques, and this did not change when radiation therapy was included as a factor. Future studies with randomization will better elucidate if there is a true functional difference in this particular patient population. CONCLUSION Patients with facial paralysis after cancer resection have the potential for improved facial symmetry by undergoing reanimation procedures. Static sling and OTM procedures are both good options and can be performed either at the time of resection or as an outpatient. Patients undergoing radiation either pre- or postoperatively can be advised that they have the potential for wound complications, but that the risk is fairly low. Reanimation procedures need not be staged, as this does not confer any excess risk to the reconstruction or increase revision or complication rates. BIBLIOGRAPHY 1. Ho AL, Scott AM, Klassen AF, Cano SJ, Pusic AL, Van Laeken N. Measuring quality of life and patient satisfaction in facial paralysis patients: a systematic review of patient-reported outcome measures. Plast Reconstr Surg 2012;130: Melvin TA, Limb CJ. Overview of facial paralysis: current concepts. Facial Plast Surg 2008;24: Leckenby JI, Harrison DH, Grobbelaar AO. Static support in the facial palsy patient: a case series of 51 patients using tensor fascia lata slings as the sole treatment for correcting the position of the mouth. J Plast Reconstr Aesthet Surg 2014;67: Chuang DC. Free tissue transfer for the treatment of facial paralysis. Facial Plast Surg 2008;24: White H, Rosenthal E. Static and dynamic repairs of facial nerve injuries. Oral Maxillofac Surg Clin North Am 2013;25: Iseli TA, Harris G, Dean NR, Iseli CE, Rosenthal EL. Outcomes of static and dynamic facial nerve repair in head and neck cancer. Laryngoscope 2010;120: Griffin GR, Abuzeid W, Vainshtein J, Kim JC. Outcomes following temporalis tendon transfer in irradiated patients. Arch Facial Plast Surg 2012;14: Skourtis ME, Weber SM, Kriet JD, Girod DA, Tsue TT, Wax MK. Immediate Gore-Tex sling suspension for management of facial paralysis in head and neck extirpative surgery. Otolaryngol Head Neck Surg 2007; 137: Boahene KD, Farrag TY, Ishii L, Byrne PJ. Minimally invasive temporalis tendon transposition. Arch Facial Plast Surg 2011;13: Revenaugh PC, Knott PD, Scharpf J, Fritz MA. Simultaneous anterolateral thigh flap and temporalis tendon transfer to optimize facial form and function after radical parotidectomy. Arch Facial Plast Surg 2012; 14: