Swallowing Function in Patients With Base of Tongue Cancers Treated With Primary Surgery and Reconstructed With a Modified Radial Forearm Free Flap

Objective  To report swallowing outcomes and biomechanical properties of the base of the tongue (BOT) and the posterior pharyngeal wall (PPW) in patients who undergo surgical reconstruction with the beavertail modification of radial forearm free flap after primary resection of BOT cancer.

Design  Prospective cohort study with a 1-year minimum follow-up performed between October 1, 2001, and August 31, 2005.

Setting  Tertiary care facility.

Patients  Patients diagnosed as having primary carcinoma of the BOT were treated with primary surgical resection and reconstruction followed by radiotherapy. Inclusion criteria were collection of videofluoroscopic swallowing study (VFSS) data before and 1 year after surgery. Forty-one patients were treated during a 5-year period, and 20 were included in the final analysis.

Interventions  Reconstruction of BOT defects with the beavertail modification of radial forearm free flap followed by postoperative radiation.

Main Outcome Measures  Aspiration score, pharyngeal residue score, and biomechanical analysis of BOT and PPW mobility were performed using images from VFSSs. Both the BOT and PPW positions were measured from 2 static bony landmarks.

Results  Of the 20 patients in the final analysis, 19 (95%) were able to swallow safely at 1 year. Mobility of the BOT after surgery was reduced in all postoperative VFSS data. Anteroposterior dimension or bulk of the BOT was preserved. No significant difference was found in PPW mobility.

Conclusions  The beavertail modification of the radial forearm free flap is a good reconstructive option after BOT cancer extirpation. The procedure preserves the bulk of the BOT after cancer treatment and maintains adequate BOT-PPW apposition. This allows structures such as the pharyngeal, oral, and suprahyoid musculature to contract and generate the necessary force to propel the food bolus through the oropharynx, resulting in a safe swallow.

Head and neck oncologists are often confronted with the difficult challenge of balancing cancer cure and patient survival with the preservation of function, cosmesis, and quality of life when deciding the patient's best treatment regimen. This is especially true for cancers of the oropharynx because this region has been intimately involved with the functions of speech, deglutition, and swallowing.

Early attempts at surgical treatment of oropharyngeal cancers invariably resulted in patients having poor functional outcomes, likely because of a lack of formal reconstruction.^1,2 Poor postoperative outcomes in these patients led many centers to experiment with organ preservation protocols using a combination of radiotherapy and chemotherapy in an attempt to preserve function. Unfortunately, early reports^3,4 of functional outcomes in patients treated with these protocols have been for the most part disappointing.

The widespread use of free tissue transfer in reconstruction has improved postoperative functional outcomes,^5-8 but a paucity of information remains regarding the swallowing outcomes in patients who have undergone oropharyngeal resections and reconstructions.^5-7,9-13 Even fewer studies^13-15 have examined the biomechanics of swallowing in the reconstructed oropharynx, including examining the specific mechanics of the base of the tongue (BOT) and the posterior pharyngeal wall (PPW) required to generate the bolus-driving force during the oropharyngeal phase of swallowing.^16,17 These studies have the significant weaknesses of small sample size and failure to control for cancer site and reconstruction technique.

The purpose of the current study was to examine the functional swallowing outcomes in a cohort of patients all diagnosed as having BOT cancer after resection and reconstruction with beavertail modification of the radial forearm free flap by (1) prospectively collecting swallowing data both before and after surgery and (2) examining the biomechanical properties of the BOT and PPW in the reconstructed oropharynx.

Forty-one consecutive patients treated for BOT cancer between October 1, 2001, and August 31, 2005, at the University of Alberta were followed up in a prospective manner through an interdisciplinary Head and Neck Reconstructive Clinic in Edmonton. Thirty-six patients were treated with primary surgery, reconstruction of the defect with a beavertail modification of the radial forearm free flap, and postoperative radiotherapy. Five patients underwent primary chemotherapy and radiotherapy. All patients participated in the functional assessment protocol, which included videofluoroscopic swallowing studies (VFSS) performed preoperatively and at 12 months postoperatively. No patients received any swallowing therapy during their follow-up visits, although all received swallowing therapy before discharge from the hospital after surgery. Each patient was given instructions for oral motor exercises to be performed at home after discharge.¹⁸ All patients received postoperative radiotherapy treatment at a single institution commencing 6 weeks postoperatively. All 36 surgical patients had 1-year follow-up and data gathered regarding clinical swallowing function, presence of gastrointestinal feeding tubes, maintenance of weight, and diet survey. All 36 surgical patients had both preoperative and postoperative VFSS times arranged. Inclusion criteria for the study were completion of both preoperative and 1-year postoperative VFSSs. Twenty patients met the inclusion criteria for this study.

All patients had primary tumors that involved the BOT, with some involvement of the lateral pharyngeal walls or soft palate. All patients had between 50% and 100% of their BOT resected. All resections were limited to the oropharynx and did not include any resection of the oral tongue or other structures in the oral cavity. Surgical access to the oropharynx was gained via lip-splitting incision and presymphyseal mandibulotomy. All patients had at least 1 intact lingual nerve, hypoglossal nerve, lingual artery, and vessels. All patients had their respective defects reconstructed with the beavertail modification of the radial forearm free flap. The proximal portion of the flap was de-epithelized, creating an accessory paddle of subcutaneous fat and fascia. This paddle was then rolled on itself and inset into the BOT defect to add bulk to the reconstructed BOT. The rest of the flap was then redraped over the neotongue and inset into the BOT.¹⁹ None of the flaps were innervated. All patients underwent tracheostomy during the initial resection and reconstruction procedure. All patients underwent successful decannulation with fully healed stoma sites before the collection of any postoperative swallowing data. Laryngeal suspension was performed for BOT resections that exceeded 50% when the suprahyoid muscles were resected bilaterally.

Before treatment, informed consent for data collection and use in clinical research was obtained from each patient. Ethics for the present study were approved by the health research ethics board at the institution where the surgical treatment of each patient would take place (University of Alberta Hospital). Relevant patient and tumor information was collected prospectively. Diagrams of the location and extent of the tissue resection were made by the primary surgeon (J.R.H. and H.S.) at surgery. The percentages of BOT and soft palate resected were also recorded at surgery.

Swallowing function was assessed by analysis of VFSS data collected preoperatively and 12 months postoperatively. The VFSS was completed at a single institution (Misericordia Community Hospital, Edmonton) in the presence of a radiologist. A pudding mixed with barium sulfate cream (Esobar; Therapex, Montreal, Quebec, Canada) in a 3:1 ratio was used for this study. The pudding was presented as a calibrated bolus on a teaspoon by the examiner. The VFSS data were recorded as digital files (KayPENTAX Digital Swallowing Workstation; KayPENTAX, Lincoln Park, New Jersey). The VFSS data were then analyzed by a qualified speech pathologist (J.R., A.S., J.Z.) for laryngeal penetration, aspiration events, and pharyngeal residue. Significant aspiration events were defined as penetration and persistence of food bolus below the level of the vocal cords, as defined by the previously validated penetration-aspiration scale.²⁰ Pharyngeal residue scores were graded as mild, moderate, and severe. Pharyngeal residue scores are used as a marker of oropharyngeal swallowing force production capability.²¹ A mild residue score indicates no difficulty in propelling the food bolus through the oropharynx; a moderate score, some impairment in bolus propulsion; and a severe score, marked impairment in propulsive force generation. Overall observations on laryngeal elevation and passage of food bolus over the BOT and through the pharynx during the swallowing sequence were recorded during the analysis of each VFSS.

Biomechanical analysis was used to assess the movement of the BOT and PPW over time. Single-frame images from the VFSS recordings of all patients were analyzed at 4 distinct phases of the swallowing sequence.²² Single frames (Figure 1) were obtained at (1) the start of the swallow (immediately after placement of food bolus on the tongue), (2) the head of bolus passing the mandibular ramus, (3) the tail of bolus passing the mandibular ramus, and (4) closure of the cricopharyngeus muscle.

The following lines and points were marked on each analyzed frame (Figure 2): (1) a line that intersected the C2 and C4 cervical vertebrae following the plane of the anterior vertebral line, (2) a line parallel to the line described in item 1 that intersects the most anterior point of the mentum, (3) a line parallel to those described in items 1 and 2 that intersects the most posterior aspect of the BOT, and (4) a line parallel to those described in items 1 and 2 that intersects the most anterior aspect of the posterior pharyngeal wall.

The BOT position and the PPW position were measured in millimeters from the line that intersected the mentum. To correct for fluoroscopic magnification, all measurements were then divided by a baseline distance, which was defined as the distance (in millimeters) between the anterior vertebral line (line 1) and the mentum (line 2). This allows for all of the measurements of BOT and PPW mobility to be reported in units that nullify the effects of fluoroscopic magnification among the different VFSSs performed during each patient's treatment course.

Measurements of BOT and PPW position were taken at the 4 previously mentioned phases of the swallowing sequence. From these data points, the following outcome measures were calculated: (1) position of the BOT at rest, which equates to tongue bulk and volume (measurements taken from start of swallow image frame); (2) position of the PPW at rest (measurements taken from the start of the swallow image frame); (3) extent of the BOT movement in the posterior direction during the swallowing sequence; and (4) extent of PPW movement in the anterior direction during the swallowing sequence.

The biomechanical analysis of BOT and PPW mobility during swallowing in the present study has 8 dependent variables across 2 different periods: before operation and 12 months after operation. A repeated-measures analysis of covariance design was used for each variable to determine the differences across time. The percentage of the BOT that was resected (range, 50%-100%) was used as a covariate because the amount resected could potentially influence the results. Using the percentage of the BOT resected as a covariate for the analysis allows for the reduction of uncontrolled variance that this may have on the BOT and PPW motion not accounted for in the design of the experiment. This method also allows for a more precise and less biased estimate of the group effect of the beavertail reconstruction technique on the biomechanics of swallowing during the data analysis. Tests of within-patient contrasts were used to compare the differences among different periods. Statistical analyses were performed using a commercially available software program (SPSS for Windows, version 14; SPSS Inc, Chicago, Illinois), and an α level of .05 was used throughout the study.

Of the 36 surgical patients, 20 had complete VFSS and swallowing data sets and were included in the final data analysis. The other 16 patients were followed up regularly through the head and neck reconstructive clinic but were unable to complete the 1-year postoperative swallow study because of various reasons. None of these patients were diagnosed as having aspiration pneumonia or other aspiration-related sequelae up to 1 year postoperatively. All 36 patients at 1 year postoperatively were taking feedings orally; 4 patients had gastrointestinal tubes in place, but only 1 patient (patient 17) was using the tube for nutritional support even though she had a safe swallow on her VFSS.

The demographics of all 41 patients with BOT cancer are given in Table 1. The study group's mean age was 54.6 years (range, 44-70 years). The group with complete data consisted of 14 men (70%) and 6 women (30%). Ten patients (50%) had stage IV disease, 8 (40%) had stage III disease, and 2 (10%) had stage II disease. The VFSS data were collected for all 20 patients at preoperative and 12-month postoperative dates.

Significant aspiration events, defined as penetration and persistence of food bolus below the level of the vocal folds on VFSS and pharyngeal residue scores for each patient preoperatively and 12 months postoperatively, are summarized in Table 2 and in Figure 3 and Figure 4.^20,21 None of the 20 patients had significant aspiration events at the preoperative VFSS. A single patient (patient 10) had a significant aspiration risk on the 12-month postoperative study. This patient, however, refused gastrointestinal tube insertion and clinically did not have any aspiration-related sequelae at subsequent follow-up visits. The 19 remaining patients (95%) had safe swallowing ability. Pharyngeal residue scores were calculated during all swallows analyzed in this study (Figure 3). Seventeen patients (85%) had mild pharyngeal residue during the preoperative VFSS, whereas the 3 remaining patients (15%) had moderate pharyngeal residue. At the 12-month postoperative VFSS, 6 patients (30%) had mild residue scores (Figure 4), 6 (30%) had moderate residue scores, and the remaining 8 (40%) had severe residue scores.

Biomechanical analysis of structural movement and BOT bulk during swallowing involved the examination of the position of the BOT and PPW at rest and at different progressive stages during the swallowing sequence. All measurements are reported as distance (in standardized units) from the mentum. This value was calculated via measuring the distance (in millimeters) from the mentum to either the most posterior position noted of the BOT (Figure 2, line B) or the most anterior position of the PPW (Figure 2, line C). These values were then divided by the distance measured (in millimeters) from the mentum to the anterior vertebral line reference point (Figure 2, distance between lines A and D). This technique accounts for and negates the effects of potential errors in distance measurements created by magnification effects on the images by patients sitting at different distances from the imaging equipment during their different VFSSs. All measurements taken from preoperative and 12-month postoperative VFSS images are summarized in Table 3. A repeated-measures analysis of covariance with percentage of BOT resected used as a covariate showed no significant differences in the position of the BOT and PPW during the swallowing sequence across the 4 periods (P values ranged from .06 to .66 for BOT and from .08 to .93 for PPW measurements). No significant difference was found between the measurements of the BOT position during rest at preoperative and 1-year postoperative studies, indicating that the BOT bulk was maintained. An anterior-directed movement of the PPW was found during the last 3 stages of the swallowing sequence. No significant difference was noted between the preoperative and postoperative VFSSs in overall movement of the PPW.

The swallowing outcomes reported in this study suggest that patients undergoing resection of oropharyngeal cancers that involve the BOT followed by reconstruction with the beavertail modified radial forearm free flap can consistently achieve efficient and safe swallowing function postoperatively. A total of 19 (95%) of the 20 patients included in this study had safe, functional swallows as determined by VFSS analysis at 12 months postoperatively.²⁰ Forty swallowing studies were analyzed in total; 20 of them were measured postoperatively. One of the postoperative studies (5%) showed significant aspiration events. The results reported herein regarding aspiration events correlate with previously published studies. Seikaly et al⁷ reported 8 instances of aspiration in 128 swallows analyzed 6 months to 1 year after treatment, representing 6% of the data. Pauloski et al²³ reported 24 instances of aspiration in 900 swallows analyzed, representing 3% of a heterogeneous group of patients who had primary closure, skin grafting, or no reconstruction of oropharyngeal defects. Logemann et al²⁴ analyzed swallowing data in patients having primary closure of oropharyngeal defects. Three significant aspiration events occurred in 158 swallows studied, representing 2% of their data. Primary closure was not feasible in any of our patients because the BOT defects were rather large, as indicated by data in Table 1; therefore, direct comparison to the results of Pauloski et al²³ and Logemann et al²⁴ is difficult. This study is different from all previously published studies that involved surgical treatments of oropharyngeal cancer in that it shows similar low aspiration rates combined with biomechanical data, which helps us explain how this type of reconstruction preserves swallowing function.

The results reported in this study are in contrast to data reported from studies that examined the posttreatment swallowing function after combined chemotherapy and radiotherapy protocols. Lazarus et al⁴ described 8 of 9 patients (89%) with significant aspiration events after combined chemotherapy and radiotherapy protocols. Kotz et al²⁵ also examined pharyngeal dysmotility after radiotherapy and intravenous hydroxyurea. Their study recorded aspiration events in 4 of 15 patients (27%). The long-term follow-up study³ by the same group showed persistence of pharyngeal dysmotility and gastrointestinal tube dependence in patients with oropharyngeal cancer treated with this chemoradiotherapy protocol. All of the patients described in our study were free of gastrointestinal tube dependence at 1 year after treatment. Most patients in our study had advanced-stage cancer (90%) in proportions similar to those reported by Lazarus et al,⁴ Kotz et al,²⁵ and Smith et al.³ The functional outcomes reported are, therefore, not related to differences in stage of cancer.

The VFSS data indicate that swallowing function is preserved in the cohort of patients discussed in this study. It appears that the anatomical features of the reconstructed oropharynx combined with the remaining pharyngeal and suprahyoid musculature enable the patients to swallow safely. Current models of swallowing suggest that during the normal swallowing sequence, the driving force pushing a food bolus through the oropharynx is the posterior movement of the BOT combined with the caudally directed sequential contraction of the pharyngeal constrictors.²² Other studies^14,16,17 have acknowledged the importance of the apposition of the BOT and PPW to allow adequate force generation by the pharyngeal constrictors to enable the caudally directed force to drive the bolus through the oropharynx.

The measurements of BOT position made in the present study indicate that in the preoperative setting there is a consistent trend in posterior-directed motion of the BOT after the initiation of the swallowing sequence, which is coupled with a consistent trend in anterograde movement by the PPW (Table 3). These findings appear to follow expected patterns predicted by current models. The 1-year posttreatment biomechanics show the following: (1) no significant change in the movement of the PPW as the swallowing sequence is followed through its different phases, (2) reduced motion of the BOT, and (3) preservation of BOT bulk because no significant difference was found in the position of the BOT at each separate phase of swallowing from the preoperative state.

The lack of major increases in the motion of the PPW throughout the various postoperative swallowing studies compared with the preoperative measurements seems to indicate that this structure is not the major compensatory mechanism that allows adequate BOT-PPW apposition. After surgery and radiotherapy, the PPW appears to act in a similar fashion to its pretreatment state, with no increase in forward- or anterior-directed movement to appose the BOT.

Biomechanical analysis indicates that although the mobility of the posterior aspect of the tongue is reduced, the bulk of the BOT remained unchanged from the preoperative state. This preservation of bulk is believed to maintain adequate BOT-PPW apposition to allow for structures such as the pharyngeal, oral, and suprahyoid musculature to contract and generate the necessary force to propel the food bolus through the oropharynx. This theory is supported by observations recorded during each VFSS, which showed that in all of the postoperative swallowing studies there was still significant laryngeal elevation noted as the head and the tail of the food bolus passed the level of the mandibular ramus. As the food bolus was propelled over the BOT, it appeared that the laryngeal elevation moved the larynx anteriorly and superiorly as is expected during the swallowing sequence. The relatively adynamic BOT appeared to act like a buttress that was observed to be protective against aspiration events. As the food bolus was forced through the oral cavity and oropharynx via muscle contraction, the bolus moved over the bulky BOT, allowing it to drop through the oropharynx and hypopharynx and, last, into the cervical esophagus, avoiding spillage into the glottic region.

The possible role of the BOT bulk in preserving swallowing function is further supported by the pharyngeal residue scores. At 1 year after treatment, 60% of the patients had adequate force production as measured by the pharyngeal residue scores; the other 40% had some impairment in force production, lowering the efficiency of driving the food bolus through the oropharynx. This, however, is in contrast to the observation that only 1 of the 5 patients with a severe pharyngeal residue score had any significant aspiration event. These findings indicate that although the muscle force driving the food bolus through the oropharynx is reduced, the patient's swallowing function is preserved. It is believed that the combination of preserved BOT-PPW apposition and the observation that the bulk of the BOT directs food away from the glottis is what preserves swallowing function in patients whose oropharyngeal musculature has been altered by surgery and radiation.

The current study has a number of acknowledged weaknesses. First, the small sample size may have resulted in reduced statistical significance. The power of the statistical analysis calculated in the present study was moderate. Second, because of the 2-dimensional nature of the biomechanical analysis, it was impossible to accurately study the mobility and movement of the lateral pharyngeal walls, especially with respect to contact with the BOT during swallowing.

Future studies would potentially use biplane videofluorography for the collection of VFSS data. This would enable examiners to determine if efficient swallowing function, as determined by pharyngeal residue score, is preserved by some large increase in lateral pharyngeal wall mobility and if increased pharyngeal residue scores are in fact indicative of reduced lateral pharyngeal wall compensation.¹³ Comparative analysis of swallowing outcomes and biomechanical analysis of BOT and PPW mobility in an age-, tumor-, and cancer stage–matched cohort of patients treated with primary chemoradiotherapy protocols would further increase our understanding of posttreatment swallowing outcomes and enhance treatment modality selection.

In conclusion, the beavertail modification of the radial forearm free flap is a good reconstructive option for BOT cancer extirpation. The procedure preserves the bulk of the BOT after cancer treatment and maintains adequate BOT-PPW apposition. This allows structures, such as the pharyngeal, oral, and suprahyoid musculature, to contract and generate the necessary force to propel the food bolus through the oropharynx, resulting in a safe swallow.

Correspondence: Hadi Seikaly, MD, Division of Otolaryngology–Head and Neck Surgery, 1E4 Walter C. MacKenzie Health Sciences Centre, 8440 112th St, Edmonton, AB T6G 2R7, Canada (hadi.seikaly@capitalhealth.ca).

Submitted for Publication: November 13, 2006; final revision received November 3, 2007; accepted November 13, 2007.

Author Contributions: All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: O’Connell, Rieger, Harris, Dziegielewski, Li, Wolfaardt, Hart, and Seikaly. Acquisition of data: O’Connell, Rieger, Dziegielewski, Zalmanowitz, Sytsanko, and Seikaly. Analysis and interpretation of data: O’Connell, Rieger, Harris, Dziegielewski, Zalmanowitz, Sytsanko, Li, and Seikaly. Drafting of the manuscript: O’Connell, Dziegielewski, and Li. Critical revision of the manuscript for important intellectual content: O’Connell, Rieger, Harris, Dziegielewski, Zalmanowitz, Sytsanko, Wolfaardt, Hart, and Seikaly. Statistical analysis: Li. Administrative, technical, and material support: O’Connell, Rieger, Harris, Dziegielewski, Zalmanowitz, Sytsanko, Wolfaardt, Hart, and Seikaly. Study supervision: Harris, Hart, and Seikaly.

Financial Disclosure: None reported.

Funding/Support: This study was partially supported by the University of Alberta Otolaryngology Donation Fund and by the University of Alberta Otolaryngology Resident Research Fund.

Previous Presentation: This study was previously presented as a poster at the American Head and Neck Society 2006 Annual Meeting and Research Workshop on the Biology, Prevention, and Treatment of Head and Neck Cancer; August 18, 2006; Chicago, Illinois.

Additional Information: This paper was awarded the American Head and Neck Society's annual Best Resident Clinical Research Award.