Left composite oropharyngeal defect and oropharyngoplasty using the Gehanno technique. A, Left composite oropharyngeal defect including the following subsites organized by tissue type (mucosa, muscle, nerve, and bone): 60% soft palate, 40% base of tongue (BOT), tonsil, anterior tonsil pillar, posterior tonsil pillar, left posterior pharyngeal wall, glossotonsillar sulcus, posterior floor of the mouth, retromolar trigone, mandibular alveolar mucosa, buccal mucosa, tensor veli palatini, middle constrictor, palatoglossus, palatopharyngeus, anterior medial pterygoid, lingual nerve, inferior alveolar nerve, anterior ramus of the mandible, body of the mandible, and the maxillary tuberosity. 1, Resected edge of the soft palate; 2, resected edge of the posterior pharyngeal wall; 3, parapharyngeal compartment; 4, partially resected medial pterygoid; 5, partially resected ramus of the mandible; 6, partially resected BOT; and 7, mandibular compartment after the resection of the body of the mandible. B, Oropharyngoplasty using the Gehanno technique of superior constrictor velopharyngoplasty and BOT mounding. 1, The superior constrictor velopharyngoplasty is achieved by closing the posterior pharyngeal wall to the nasal side of the soft palate; in a second layer the superior constrictor is closed to the tensor veli palatini; 2, the BOT mounding is achieved by primary closure with several large 2.0 monofilament sutures; 3, there is a large volume defect in the parapharyngeal and mandibular compartments, which will be filled by revascularized tissue; 4, a gap is shown at the superior edge of the velopharyngoplasty, which is closed with the revascularized free tissue transfer; and 5, the feeding tube should be inserted prior to creation of the velopharyngoplasty.
Lateral view of the oropharynx, nasopharynx, oral cavity, and the laryngeal inlet. A, Tongue base mounding is completed as shown. The preparatory incision to separate and mobilize the oral side of the palate and the nasal side of the palate is shown prior to performing the superior constrictor velopharyngoplasty. 1, Oral tongue; 2, soft palate with cross-section of tensor veli palatine; 3, residual base of the tongue, which has been closed primarily to the oral tongue; 4, incision to separate oral and nasal sides of the soft palate; 5, posterior pharyngeal wall advancement rotation flap; 6, superior pharyngeal constrictor; 7, occlusal line; 8, tip of the epiglottis; 9, vallecula; and 10, inlet of larynx. B, Velopharyngoplasty has been performed. The nasal side of the of the soft palate extending along the contralateral posterior tonsillar pillar is sutured to the posterior pharyngeal wall that bordered the resection margin. The goal is to create a myomucosal tube with the residual oropharyngeal tissues, which extends 2.5 to 3.0 cm below the occlusal line. Very little soft palate dissection is required. The velopharyngoplasty is customarily started above the occlusal line as shown. 1, Oral palate mucosa; 2, superior aspect (inlet) of velopharyngoplasty; 3, soft palate incision to mobilize the oral and nasal surfaces of the soft palate; 4, oral side of soft palate; 5, superior pharyngeal constrictor, which has been advanced with its overlying posterior pharyngeal mucosa to form the velopharyngoplasty; 6, closure line between the nasal side of the soft palate and the posterior pharyngeal wall; 7, occlusal line; 8, tip of the epiglottis; 9, inferior aspect (outlet) of velopharyngoplasty; and 10, inlet of larynx.
Revascularized free tissue L-shaped transfer template. BOT indicates base of the tongue; NP, nasal palate; and RMT, retromolar trigone.
Nine-year follow-up of a 55-year-old man who underwent resection and reconstruction for a second primary T4N1M0 squamous cell carcinoma of the tonsil. A, An L-shaped template of the anatomic subunits for the revascularized reconstruction. The crosshatched portion is de-epithelialized and is used for the restoration of the parapharyngeal and mandibular compartments. The template is applied to the patient's left arm. The layout is angled to optimize the use of the brachioradialis fat pad and to bring the donor vessels over the carotid space. ALV indicates alveolar mucosa; BOT, base of the tongue; BUC, buccal mucosa; MAX, maxillary tuberosity; NP, nasal palate; OP, oral palate; RMT, retromolar trigone; TON, tonsil. B, Intraoral view. The field of focus of the photograph is on the reconstructed 75% soft palate defect. Note that the tongue base mound remains high in the patient's oropharynx in a position that facilitates good contact with the reconstructed palate. The revascularized free tissue transfer is on the patient's left palate and extends though the retromolar trigone, along the reconstruction plate to the anterior floor of the mouth. Note also that despite a 75% palate defect and a 33% base of tongue defect, the inferior margin of the soft palate cannot be visualized owing to the length of the velopharyngoplasty.
Greater than 12-month median scores on the Head and Neck Speech and Swallowing Assessment based on defect group. “Eating in Public” and “Understandability of Speech” are validated scales from the Performance of Speech and Swallowing survey instrument. All questions use a 5- to 6-point Likert scale. Patients in group 2 showed a trend to lower swallowing scores, but there was no statistically significant difference. This reconstructive approach was effective in rehabilitating speech independent of defect size. BOT indicates base of the tongue.
Twelve-month median scores on the University of Michigan Head and Neck Quality of Life (UM HNQOL) scale based on defect group. Although the trend is worse in the larger defect group, there is no statistically significant difference. BOT indicates base of the tongue.
Long-term median scores on the University of Michigan Head and Neck Quality of Life (UM HNQOL) scale based on defect group. BOT indicates base of the tongue. *All domains were significantly different with the exception of the speech domain.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Chepeha DB, Sacco AG, Erickson VR, et al. Oropharyngoplasty With Template-Based Reconstruction of Oropharynx Defects. Arch Otolaryngol Head Neck Surg. 2009;135(9):887–894. doi:10.1001/archoto.2009.130
To determine if oropharyngoplasty using a Gehanno technique of superior constrictor velopharyngoplasty, base of tongue mounding, and primary hypopharyngeal closure in combination with template-based revascularized free tissue transfer is effective for reconstruction of the oropharyngeal defect.
Prospective case series.
Tertiary care academic medical center.
The study population comprised 25 patients (21 men and 4 women; mean age, 55.3 years) presenting from January 1998 to January 2001 with oropharyngeal squamous cell carcinoma. A comparison was performed based on the percentage of resection of the soft palate (group 1, ≤50% palate; group 2, >50% palate).
Of the 25 patients, 24 (96%) received radiotherapy. The donor sites were radial forearm for 23 of 25 patients (92%) and lateral arm for 2 of 25 patients (8%). The mean area was 92 cm2 (range, 25-150 cm2), and the mean length of the velopharyngoplasty component of the oropharyngoplasty was 2.15 cm (range, 1-3 cm).
Main Outcome Measures
Gastrostomy tube dependence, major and minor complications, time to oral intake, speech and swallowing assessment, and quality-of-life assessment.
Of the 25 patients, 2 (8%) remain gastrostomy dependent; 6 (24%) developed major complications; and 7 (28%) developed minor complications. Speech in both groups 1 and 2 was considered understandable most of the time, with occasional repetition. The group 1 patient with a median assessment score could eat a solid diet without restriction of place or person, whereas the group 2 patient with a median assessment score could eat a soft, moist diet with selected persons in selected places.
Integration of oropharyngoplasty with template-based revascularized free tissue transfer produced speech results that were independent of palate defect size, and swallow function test results were similar to other published reconstructive techniques.
The oropharynx is a complex neuromuscular organ that controls velopharyngeal closure and initiation of involuntary swallow and is essential for airway protection. Currently, there are no effective techniques for replacing this complex neuromuscular organ. As a result, reconstructive efforts have focused on the maximal use of local tissue to maintain the critical functions of the oropharynx. The reconstruction of tonsil-palate defects is necessary if either oropharyngeal obliteration, velopharyngeal competence, or loss of contact of the base of the tongue (BOT) to the posterior pharyngeal wall results in impaired swallowing. The residual defect can be reconstructed with skin grafts, regional flaps, or revascularized tissue transplantation. The residual defect can also incorporate a prosthetic velopharyngeal obturator that has to be retained by a maxillary dental prosthesis.
The local tissue reconstructive options include the superiorly based posterior pharyngeal flap, superior constrictor velopharyngoplasty, and the superior-constrictor advancement-rotation flap (SCARF) flap.1-6 The superiorly based pharyngeal flap was initially described by Bardenheuer1 in 1892 and was further developed by Sanvenero-Rosselli2 in 1935. Most recently, Brown et al3 demonstrated improved speech and swallowing results when using a superiorly based pharyngeal flap. This flap was integrated into the free flap inset but was not used to formally reconstruct the nasopharyngeal sphincter. Gehanno et al,4 closely followed by Kavanagh,5 described a velopharyngoplasty with a superior constrictor flap. This approach involved reconstructing a nasopharyngeal sphincter by closing the superior constrictor to the nasopharyngeal surface of the soft palate. In contrast, Kimata et al7 compared small, medium, and large palate defects as well as 4 different approaches to palate reconstruction. These approaches included a simple patch technique, a palatal adhesion technique, a folded flap technique, and the superior constrictor flap (Gehanno technique). This study found that the superior constrictor velopharyngoplasty was associated with the lowest incidence of palatal dehiscence despite its use in medium and large defects.
An excellent alternative for large defects is the folded flap and/or the adhesion (folded flap/adhesion) approaches.8,9 The potential limitation of the folded flap and/or the adhesion approach is that a circular sphincter is not constructed, which may result in an adynamic area in the velopharyngeal port. In addition, the adhesion technique has a higher incidence of dehiscence than the superiorly based pharyngeal flap or the Gehanno technique.7 We therefore developed an oropharyngoplasty technique that includes an extended Gehanno velopharyngoplasty combined with primary closure of the hypopharynx and BOT in concert with a template-based revascularized free tissue transfer. Our goal was to optimize speech and swallowing with a tonsil-palate reconstruction that maximized the mobility and function of local oropharyngeal tissue. To determine the effectiveness of this approach, we evaluated the speech, swallowing, and quality-of-life outcomes, accounting for the size of the soft palate defect, and compared these results with the available literature.
This prospective case series included 37 patients with major ablative defects in the oropharynx reconstructed with an oropharyngoplasty technique, which included an extended Gehanno velopharyngoplasty combined with primary closure of the hypopharynx and BOT in concert with an L-shaped template-based revascularized free tissue transfer. The patients were treated by surgeons in the microvascular program of the Department of Otolaryngology–Head and Neck Surgery at the University of Michigan Health System from January 1998 to January 2001.
Patients were eligible if the surgical defect included excision of the soft palate, tonsil, and BOT; reconstruction used an oropharyngoplasty technique that includes an extended Gehanno velopharyngoplasty combined with primary closure of the hypopharynx and BOT in concert with an L-shaped template-based revascularized free tissue transfer (Figures 1, 2, 3, and 4). Twelve patients were excluded: 1 had insufficient data, 3 had rapid recurrence prior to assessment, and 8 had more extensive defects that included a laryngectomy, oral tongue, or hypopharyngeal defect where primary closure was not possible. Based on these eligibility criteria, 25 of the 37 patients were eligible to participate.
There were 21 men and 4 women, with a mean age of 55.3 years (range, 36-76 years). Mean follow-up time was 60.7 months (range, 2-115 months). Of the 25 patients, 11 (44%) are presently alive and free of disease. Five patients are dead of disease, 3 died of other disease, 5 died of intercurrent illness, and 1 is lost to follow-up. All 25 patients had squamous cell carcinoma. The radial forearm was the donor site in 23 patients, and the lateral arm was used in 2 patients (Figure 4A). All 25 patients underwent neck dissection. Of the 25 patients, 24 (96%) underwent radiation therapy; 1 received preoperative radiotherapy; 23 received postoperative radiotherapy; and 1 refused radiotherapy. Three of the patients who underwent radiation therapy also underwent concomitant postoperative chemotherapy and radiation therapy.
The tonsil palate defect was stratified into 2 groups based on the percentage of resection of the soft palate and BOT (Table). Group 1 included 10 patients with a 50% or less soft palate and 50% or less BOT defect. Group 2 included 15 patients with a greater than 50% soft palate and 50% or less BOT defect. Of the 25 patients, 6 required a mandibulectomy: 1 had a rim mandibulectomy and 5 had composite segmental resection, which was reconstructed with a 2.7-mm bridging reconstruction plate. The mean flap area was 92 cm2 (range, 25-150 cm2). Velopharyngoplasty tube length was documented in 17 of the 25 patients (68%), with a mean length of 2.15 cm (range, 1-3 cm). The revascularized free tissue transfers all had large de-epithelialized components to restore the parapharyngeal and mandibular compartments (Figure 4A).
Our approach for oropharyngeal reconstruction is to optimize the function of the remaining oropharyngeal tissues. Optimally, the patient should be able to maintain the “suction pump” function that is critical to airway protection.10-13 The suction pump is a 2-part biomechanical phenomenon, in which the oropharynx generates a positive propulsive force and the hypopharynx generates a negative pressure through laryngeal elevation and relaxation of the cricopharyngeus. The oropharynx needs a mobile BOT, a competent velopharyngeal segment, and contact of all mucosal surfaces to generate a positive pressure. Therefore, the goals of tonsil-palate reconstruction are as follows:
Obliteration of the oropharynx—obliteration is achieved when all mucosal surfaces are in contact during a swallow. Obliteration of the oropharynx is important because it decreases the likelihood of food or secretions being lost in a potential “dead space” in the oropharynx by bringing the revascularized free tissue transfer in contact with the remaining native mucosa. The remaining sensate, native mucosa is critical for triggering the swallow reflex and preventing aspiration.
Maintenance of nasopharyngeal competence.
Maintenance of BOT mobility—the BOT should be able to fully contact the posterior and lateral pharyngeal walls.
To achieve these goals in tonsil-palate reconstruction we are guided by the following principles: (1) functional primary closure of the remaining native pharyngeal tissue; (2) restoration of the nasopharyngeal sphincter with native mucosal tissues; (3) flap design and inset must allow for the anterior and posterior excursion of the BOT; and (4) the volume loss associated with the mandibular, parapharyngeal, and BOT resections compartments should be specifically replaced.
The nasopharyngeal side of the soft palate remnant was closed to the cut edge of the posterior pharyngeal wall (Figure 1 and Figure 2). The closure would bring the superior constrictor into apposition with the tensor veli palatini. The superior constrictor, which was deep to the posterior pharyngeal wall, was mobilized by dissecting it off the prevertebral fascia. This approach preserved the lateral neurovascular supply to the oropharyngeal mucosa. The velopharyngoplasty was designed to form a tube that extended 2.5 to 3.0 cm below the occlusal line. This tube usually extended to the level of the tip of the epiglottis. The goal was a sensate neuromuscular tube (Figure 1B and Figure 2B). When the resection extended to include the hard palate, the most superior portion of the tube could not be closed. In this situation, a portion of the revascularized free tissue transfer was tacked in place to facilitate closure (Figure 3). Next, if necessary, the hypopharynx was closed primarily up to the level of the vallecula (Figure 1B). Then, as necessary, the BOT mound was created. This was achieved by primary closure of the posterior aspect of the BOT to the lateral aspect of the BOT (Figure 1B and Figure 2B). The objective was to obtain 2.5 cm of height from the reconstructed glossotonsillar sulcus to the most superior aspect of the BOT. This maneuver separated the tongue from the revascularized free tissue transfer, which should improve tongue mobility as well as contact with the lateral and posterior pharyngeal walls. If the BOT was not flexible enough or the BOT defect was too large, a tab was created on the flap to reconstitute the BOT (Figure 3). These 3 maneuvers constituted the oropharyngoplasty (velopharyngoplasty, hypopharyngeal closure, and BOT mounding). The goal was to optimize the contact of the remaining oropharyngeal tissue by reconstituting the velopharyngeal sphincter and optimizing the mobility of the BOT. The primary closure of the local tissues enlarged the width of the defect that the revascularized free tissue transfer needed to resurface.
Next, the L-shaped flap template was designed by measuring the edges of the defect (Figure 3 and Figure 4A). The height of the “L” was determined by measuring from the most inferior aspect of the posterior pharyngeal defect to the most superior aspect of the palate defect, while following the natural curve from the posterior pharyngeal wall up onto the oral palate. This was usually 8 to 10 cm and is side “A” in Figure 3. Next, the base of the “L” was determined by measuring from the most inferior (and contralateral) aspect of the pharyngeal defect around the posterior pharyngeal wall, over the glossotonsillar sulcus to the most anterior aspect of the defect on the mandibular alveolus and floor of the mouth. This was usually 8 to 10 cm and is side “B” in Figure 3. Next, the width of the vertical part of the “L” was determined by measuring from the most contralateral side of the oral palate defect across the superior aspect of the defect to the maxillary buccal-alveolar sulcus. This was usually 4 to 5 cm and is side “C” in Figure 3. Next, the vertical part of the inside of the “L” was verified by measuring the superior-lateral aspect of the defect in the maxillary buccal-alveolar sulcus to the area of the base of the retromolar trigone. This was usually 5 to 7 cm and is side “D” in Figure 3. Next, the width of the base of the “L” was determined by measuring from the floor of the mouth across the most anterior aspect of the defect to the mandibular buccal-alveolar sulcus. This was usually 3 to 4 cm and is side “E” in Figure 3.
There were 3 additional subunits (tabs) that were added, if required, to the “L” template. The first was a tab for the retromolar trigone and buccal mucosa and was added in all the cases. The second was for a nasal palate defect and was added onto side “A,” de-epithelialized, folded over, and tacked into the nasal palate defect. The third was for a BOT defect that could not be closed primarily and was added to side “B.” The tab was usually needed for the 50% or greater BOT defects and was folded up onto the BOT, which recreated the glossotonsillar sulcus and resurfaced the BOT defect (Figure 3).
Fat from the brachioradialis fat pad was added along side “B” and was used to obliterate the volume lost in the parapharyngeal space, the BOT, and the mandible as appropriate. This fat could be easily tacked in at the completion of the inset.
The inset of the revascularized tissue transfer was performed in a sequential fashion similar to the method that was used to measure the template. The corner of the “L” where sides “A” and “B” meet is the most critical suture. Side “A” was closed first, and then sides “B” and “C” were draped into the defect to ensure a good fit. Then side “C” followed by side “B” were closed. The inside of the “L” was draped into the defect, sized, trimmed, and closed (Figure 4B).
Outcome measures included gastrostomy tube (G-tube) dependence, major and minor complications, time to oral intake, the Head and Neck Speech and Swallowing Assessment score, and the University of Michigan Head and Neck Quality of Life (UM HNQOL) score. Minor complications were defined as those that required treatment such as packing, drainage, medication, or surgical procedures that did not involve the movement or transfer of tissue. Major complications were defined as those that required a procedure in the operating theater with movement of tissue or resulted in death.
The Head and Neck Speech and Swallowing Assessment is a disease-specific, 6-item administered questionnaire to evaluate posttreatment speech and swallowing ability. Two of the questions are from the 3-item performance status scale.14 This assessment was performed at greater than 12 months after surgery.
The UM HNQOL15 instrument has 5 domains: speech, eating, emotion, pain, and bother. This instrument was completed by patients at the same survey intervals as the Head and Neck Speech and Swallowing Assessment.
Univariate data was tabulated on G-tube dependence, complications, assessment of speech and swallowing, and HNQOL. Mann-Whitney testing was used to compare medians within and between the 2 defect groups for the HNQOL assessment. All data were maintained in a Filemaker Pro 6.0 relational database (Claris Corp, Santa Clara, California) and kept on a server designed to protect patient confidentiality at the University of Michigan. The data were exported to a Microsoft Excel 2002 spreadsheet (Microsoft Corp, Redmond, Washington) for analysis. This is an institutional review board–approved study.
Of the 25 patients, 6 (24%) developed a major complication, which included flap loss (n = 2), fistula (n = 1), plate fracture (n = 2), and plate exposure (n = 1). Seven patients (28%) developed a minor complication, which included wound infection (n = 1), seroma (n = 2), wound abscess (n = 3), infected hematoma leading to flap dehiscence (n = 1), flap dehiscence (n = 1), fistula (n = 1), and late nasopharyngeal port stenosis (n = 2). Some of the major complications were unusual and merit further discussion. One of the 2 flap losses occurred due to the inability to reperfuse a lateral arm free flap, and a latissimus flap was used for salvage. The second flap loss occurred in a patient with underlying Raynaud phenomenon. The patient developed Raynaud phenomenon in the early postoperative period when his hospital room was cooled. The Raynaud phenomenon was associated with diminished flap perfusion and subsequent flap loss. The patient with the “major” fistula complication had a persistent orocutaneous fistula secondary to mandibular nonunion that was reconstructed with a radial forearm free flap. One of the 2 plate fractures in this series was secondary to a fall 3 years after her initial surgery; she underwent reconstruction with a fibular osteocutaneous free flap. One of the patients with nasopharyngeal port stenosis has diffuse severe postradiation changes including trismus and esophageal stenosis. Despite 2 operative attempts at transoral dilation, he remains G-tube dependent.
Of the 25 patients, 2 (8%) remain G-tube dependent. The median number of days required for patients to resume oral intake in groups 1 and 2 were 25 (range, 15-237) days and 54 (range, 10-680) days, respectively. As previously noted, group 1 patients had a 50% or less soft palate and 50% or less BOT defect, whereas group 2 patients had a greater than 50% soft palate and 50% or less BOT defect. The patient who resumed oral intake after 680 days was G-tube dependent secondary to aspiration. In the second postoperative year, after swallowing therapy and appropriate strategies, the G-tube was permanently removed.
The Head and Neck Speech and Swallowing Assessment was included as a “functional measure.” This assessment was completed in all patients who were alive and free of disease. The mean time for completion of the assessment was 28 months (range, 11-75 months). One patient in group 1 refused to participate. This resulted in 7 of 10 patients in group 1 and 9 of 15 patients in group 2. There was no statistically significant difference for speech or swallowing between group 1 and group 2, although the median score for “nutritional mode,” “range of solids,” and “eating in public” was lower in group 2. The differences between the groups for swallowing function are related to the statistical distribution of the outcomes. For group 1, the swallowing outcomes are not normally distributed and are skewed toward the higher scores. In contrast, the scores for group 2 are distributed more evenly across the Likert scale. When speaking, patients in both groups 1 and 2 required only occasional repetition and were confident speaking in any context. Therefore, the trend in these data suggests that this reconstruction performs well for the rehabilitation of speech and is independent of palate defect size.
In contrast, for swallowing outcomes this reconstructive approach performs well for palate defects of 50% or less but less well for palate defects greater than 50%. For “range of solids,” more than half of the patients (9 of 16) scored 5 of 6 or 6 of 6, which means that these patients had no dietary exclusions or minor specific exclusions such as bread crumbs. When separated by group, the “range of solids” median score was 4 of 6 for group 1 and 3 of 6 for group 2. This means that the group 1 patient with a median score could consume a variety of solids but needed liquid chasers, whereas the group 2 patient with a median score could consume a minced, moist, and soft diet. The food choices that the patients were able to make were also reflected in the “eating in public” scale. For “eating in public,” more than half the patients (9 of 16) scored 4 of 5 or 5 of 5, which means that these patients had no restrictions as to location but did have restrictions with certain difficult foods. When separated by group, the “eating in public” median score was 5 of 5 for group 1 and 3 of 5 for group 2. This means that the patient in group 1 with a median score had no restriction, whereas the patient in group 2 with a median score eats only in selected places with selected persons.
The HNQOL data from the 12-month posttreatment interval was collected on 7 of 10 patients in group 1 and 7 of 15 patients in group 2. Higher scores denote fewer problems in each domain (Figure 5). The greatest difference between the 2 groups was in the eating domain, and this correlates with the findings from the Head and Neck Speech and Swallowing Assessment; however, none of these scores achieved statistical significance (Figure 6). In contrast, long-term HNQOL data did show statistically significant differences. Long-term HNQOL data were collected between 25.5 and 71.0 months after surgery (mean, 43.5 months). This yielded some interesting results that revealed improvement with group 1 scores vs deterioration with group 2 scores across all HNQOL domains as the duration of follow-up was increased. The group 1 median scores increased 12.5 points (range, 2.1-25.0) and the group 2 median scores decreased 25 points (range, 0-46.9). As a result, the HNQOL scores from the long-term surveys demonstrate statistically significant differences between groups 1 and 2 in all domains except for the speech domain (Figure 7). With the clear differences seen between group 1 and group 2, within-group testing was performed comparing early vs late HNQOL scores despite the small sample size. Within group 1, there was no statistically significant improvement, but within group 2 there was statistically significant deterioration in the pain domain (64.3/100 vs 32.3/100; P = .01).
Integration of local oropharyngoplasty with template-based revascularized free tissue transfer is an effective approach for the reconstruction of large oropharyngeal defects that involve the palate, the pharyngeal wall and 50% or less of the BOT. The G-tube dependence rate was 8%, which compares favorably with folded flap/adhesion approaches.8,9 Our assessments do not suffer from reporter bias because all patients who underwent this reconstruction and met eligibility criteria during accrual period were included in this study.
This reconstruction maintained speech function as well or better than other published reports, particularly in the greater than 50% palate defect group. “Rehabilitation of speech” showed that patients spoke normally or required only occasional repetition and “speaking in public” showed that patients were confident speaking in any context. These results were independent of the size of the palate defect. It is unfortunate that nasometry measurements, which could have validated these speech results, were not obtained consistently. It is our opinion that attention to maximizing the length of the palate tube and mounding of the tongue were important in optimizing speech function.
The swallowing results with this approach are as good as the results published with the folded flap/adhesion approach or the posterior pharyngeal flap approach. In general, the literature shows that patients with larger palate defects perform more poorly than patients with smaller defects. The exception to these findings was published by Seikaly et al9 who used a folded flap/adhesion technique and showed that in his patients, swallowing results were independent of defect size. There are not enough data in the literature to determine if the extended superior constrictor velopharyngoplasty is a superior approach to the folded flap/adhesion approach or the posterior pharyngeal flap approach. What is consistent in the literature, however, is the importance of narrowing the nasopharyngeal port and the use of local tissue to facilitate these reconstructive maneuvers. Brown et al3 showed that a superiorly based posterior pharyngeal flap improved swallowing in 75% to 100% palate defects and Seikaly et al9 commented that he had modified their adhesion approach to ensure that the adhesion was done to the native posterior pharyngeal mucosa and not to the portion of the pharyngeal wall reconstructed by the radial forearm free flap. The extended superior constrictor velopharyngoplasty also meets these criteria and may possibly be a better adhesion technique as the length of the adhesion is longer; it creates a sensate sphincter and does not require a strip of posterior pharyngeal wall. In addition, Kimata et al7 found that the rate of postoperative palatal dehiscence was the lowest with the Gehanno technique (superior constrictor velopharyngoplasty), suggesting that this approach is one of the more reliable closure techniques. A comparison will have to be made that matches defect and makes use of the same metrics to determine which of these techniques are superior.
The oropharyngoplasty using a superior constrictor velopharyngoplasty, BOT mounding, and primary hypopharyngeal closure in combination with template-based revascularized free tissue transfer is an extension of previous reconstructive efforts which combine local tissue with vascularized tissue. The approach to the reconstruction of the palatal sphincter is most similar to the Gehanno superior constrictor velopharyngoplasty. The technique was modified so that the palate tube extended 2 to 3 cm inferior to the occlusal line. Not only does this lengthen the palatal sphincter but it also places the nasopharyngeal port in a more inferior position in the oropharynx. This maneuver should also improve the competence of the sphincter. To develop a palate tube of adequate length, particularly in near total palate defects, it is necessary to make an incision along the free edge of the remnant palate extending inferiorly and just anterior to the palatopharyngeus to a level 2 to 3 cm below the occlusal line. Creation of the palate tube, BOT mound, and closure of the hypopharynx resulted in substantial reduction in the size of the native oropharynx. Although there were some concerns about “micro”oropharynx, just as with microstomia associated with lip reconstruction, a smaller sensate functional organ is more effective than an insensate, anatomically correct, nonfunctional organ.
To assess the patients' perception of the stability of the reconstruction, long-term HNQOL data was collected. Our findings showed that emotion, pain, bother, and swallowing scores deteriorated in the greater than 50% palate defect group and improved in the 50% or less defect group. The speech domain was stable, suggesting that the reconstruction, at least as it relates to the creation of the palate tube, also remains stable. The deterioration in the larger defect group is likely due to issues related to treatment, rehabilitation, and changes in the local tissue over time. The larger defect group had deeper resections, more extensive 100% isodose radiation fields, more challenges with rehabilitation, and more atrophy of the target tissues, which includes both the reconstruction and the local native tissue.
At present, combined chemotherapy and radiation therapy is the treatment of choice for patients with oropharyngeal carcinoma. As further work is done with molecular markers, it is likely that a subgroup of patients will benefit from a primary surgical approach. With the oropharyngoplasty presented approach presented, patients undergoing primary surgery have an opportunity to retain functional speech and swallowing. It is likely that this oropharyngoplasty approach will also be used in the salvage setting for patients in whom primary chemotherapy and radiation therapy fail. As the tissues are less pliable after chemotherapy and radiation therapy, more dissection may be required to form the velopharyngoplasty. In addition, if the neuromuscular function has been affected by the prior treatment, swallowing function may be further impaired. Nevertheless, careful use of local tissue to optimize its residual function and careful flap design with attention to the reconstructed volume should optimize the reconstructive outcome available to the patient.
In conclusion, oropharyngoplasty using a modified Gehanno technique of superior constrictor velopharyngoplasty, BOT mounding, and primary hypopharyngeal closure in combination with template-based revascularized free tissue transfer is effective for reconstruction of the oropharyngeal defect. The speech results were particularly good, as there were no statistically significant differences seen with the size of the palate defect. Swallowing results were similar to folded flap/adhesion techniques and posterior pharyngeal flap techniques. This new approach offers some intuitive advantages, which may make it a more reliable alternative to the folded flap/adhesion or posterior pharyngeal flap techniques.
Correspondence: Douglas B. Chepeha, MD, MSPH, Department of Otolaryngology–Head and Neck Surgery, University of Michigan Health System, 1904 Taubman Center, 1500 E Medical Center Dr, Ann Arbor, MI 48109-0312 (firstname.lastname@example.org).
Submitted for Publication: May 20, 2008; final revision received October 14, 2008; accepted October 29, 2008.
Author Contributions: Dr Chepeha 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: Chepeha and Erickson. Acquisition of data: Chepeha, Sacco, Erickson, Lyden, Haxer, Moyer, Teknos, Prince, and Wolf. Analysis and interpretation of data: Chepeha, Sacco, Prince, Eisbruch, and Bradford. Drafting of the manuscript: Chepeha and Wolf. Critical revision of the manuscript for important intellectual content: Chepeha, Erickson, Lyden, Haxer, Moyer, Teknos, Prince, Eisbruch, and Bradford. Statistical analysis: Chepeha. Obtained funding: Chepeha and Wolf. Administrative, technical, and material support: Chepeha, Sacco, Erickson, Moyer, Teknos, Prince, and Eisbruch. Study supervision: Chepeha, Teknos, and Prince.
Financial Disclosure: KLS Martin supported a research coordinator (Ms Sacco) in the Department of Otolaryngology–Head and Neck Surgery, University of Michigan Health System, during a portion of this study.
This article was corrected online for typographical errors on 9/21/2009.