TRANSORAL ROBOTIC SURGERY FOR THE MANAGEMENT OF HEAD AND NECK CANCER: A PRELIMINARY EXPERIENCE

Transcription

1 ORIGINAL ARTICLE TRANSORAL ROBOTIC SURGERY FOR THE MANAGEMENT OF HEAD AND NECK CANCER: A PRELIMINARY EXPERIENCE Eric M. Genden, MD, Shaun Desai, BA, Chih-Kwang Sung, MD, MS Department of Otolaryngology, Head and Neck Surgery and Immunobiology, Head and Neck Cancer Center, Mount Sinai School of Medicine, One Gustave Levy Place, New York, New York eric.genden@mssm.edu Accepted 18 July 2008 Published online 28 October 2008 in Wiley InterScience ( DOI: /hed Abstract: Background. The aim of this prospective study was to determine the technical feasibility, safety, and efficacy of transoral robotic surgery (TORS) for a variety of malignant head and neck lesions. Methods. From April 2007 to November 2007, 20 patients were enrolled in an institutional review board approved prospective trial using the davinci surgical robot. Inclusion criteria for the study consisted of adults with early head and neck cancer involving the oral cavity, oropharynx, hypopharynx, and larynx. Results. Twenty patients were included in this study. In 2 cases, access to the tumor was inadequate and the procedure was terminated. In all 18 cases, negative resection margins were achieved. Intraoral reconstruction was performed in 8 patients. Fifteen of 18 patients underwent concomitant unilateral (n 5 10) or bilateral (n 5 5) selective neck dissections. None of the patients required tracheotomy and there were no intraoperative or postoperative complications. The average setup time was 54.6 minutes (range, minutes), with a precipitous decrease in the setup time as the study progressed. Conclusion. TORS is a safe, feasible, and minimally invasive alternative to classic open surgery or endoscopic transoral laser surgery in patients with early cancer of the head and neck. With increasing experience, surgical setup as well as operative time will continue to decrease. VC 2008 Wiley Periodicals, Inc. Head Neck 31: , 2009 Keywords: robot; surgery; transoral; laser; minimally invasive Correspondence to: E. M. Genden VC 2008 Wiley Periodicals, Inc. Since the development of the first surgical robot in 1985 by Kwoh et al, 1 there has been a slow but steady development of robotic technology designed to improve the accuracy and efficiency of surgery. Robotic technology has been applied to urologic surgery, 2 orthopedic surgery, 3 general surgery, 4 cardiothoracic surgery, 5 and most recently, otolaryngology. 6 8 Using the davinci Surgical System (Intuitive Surgical, Sunnyvale, CA), O Malley and Weinstein et al have documented the potential value of transoral robotic surgery (TORS) for the management of tumors of the upper aerodigestive track. 6,7,9 Additionally, they reported a safety profile for TORS demonstrating its safety for use in human patients. 10 This work hints at the potential for future developments for robotic surgery in otolaryngology; however, the ideal applications and limitations of TORS have not yet been elucidated. This prospective study was designed to evaluate the technical feasibility, safety, and efficacy of TORS for a variety of malignant head and neck lesions and review the learning curve associated with introducing a TORS program into a tertiary care medical center. Transoral Robotic Surgery in Head and Neck Cancer HEAD & NECK DOI /hed March

2 FIGURE 1. Positioning of the robotic camera and instrument guidance trochars using the Dingman retractor. [Color figure can be viewed in the online issue, which is available at www. interscience.wiley.com.]. PATIENTS AND METHODS Patients. From April 2007 to November 2007, a total of 20 patients were identified as candidates for TORS in a Mount Sinai Hospital Institutional Review Board approved pilot study designed to evaluate the application of TORS for the management of malignant disease of the upper aerodigestive system. Inclusion criteria for the study consisted of adults 18 years old and over, regardless of sex, race, or ethnicity, with early head and neck cancer involving the oral cavity, oropharynx, hypopharynx, and larynx. Informed consent was obtained from all patients. Patients were evaluated for age, sex, primary site of the tumor, pathology, and stage of disease. Each case was analyzed for time to setup the instrumentation, surgical time, access to the lesion, complications, ability to achieve tumor-negative margins, management of neck disease, method of reconstruction, postoperative diet, speech, and adjuvant therapy. Surgeon Training. All surgeries were performed by the principal investigator (E.M.G.) with the aid of a scrubbed assistant. Surgeon training consisted of a training course performed on a series of cadaver and porcine models at Intuitive Surgical and Hackensack University Medical Center, Hackensack, NJ. Following training, the principal investigator (PI) enrolled patients for a pilot trial. This study represents evaluation of only those patients enrolled in the trial. Robot Configuration, Operating Room Setup, and Operative Procedure. The davinci Surgical System consists of a surgeon s console, a surgical cart, a manipulator unit with 2 laterally placed instrument arms, and a centrally located endoscopic arm holding the 3-dimensional (3D) camera. As previously described by O Malley et al, 7 the patient was placed in the supine position on an operating room table. After standard endotracheal intubation, the davinci robot was positioned at a 308 to 458 angle to the operating room table. Oral cavity retraction was achieved using the Crowe Davis retractor, Dingman retractor, or the FK retractor (Gyrus Company, Maple Grove, MN) as described previously (Figure 1). 7 In each case, the retractor was suspendedfromaretractionarmfastenedseparately to the operating room bed. The 08 or 308 high-magnification 3D camera was inserted into the oral cavity followed by positioning of the dual robotic arms. Surgical resections and reconstructions were performed using 8-mm robotic instruments including needle drivers, bioploar forceps, Maryland forceps, and a 200-lm spot-size flexible carbon dioxide (CO 2 ) LASER (OmniGuide Company, Cambridge, MA) with a 20 W NovaPulse LASER (Lumenis, Santa Clara, CA). RESULTS Sixteen men and 4 women with an average age of years (range, years) were included in this study (Table 1). In 2 cases, access to the tumor Table 1. Patient demographics and pathology reports. Case Age Sex Site Pathology Path stage (TNM) Outcome 1 56 F Tonsil SCC T1 N0 M0 Success 2 49 M Tonsil SCC T1 N0 M0 Success 3 47 M Tonsil SCC T1 N0 M0 Success 4 66 M Tonsil SCC T1 N1 M0 Success 5 38 M BOT SCC T1 N1 M0 Success 6 78 M Postpharynx SCC T2 N2a M0 Success 7 48 F Palate SCC T2 N0 M0 Success 8 51 M BOT CR cell T2 N0 M0 Success 9 56 F Epiglottis SCC T1 N1 M0 Success M AE fold SCC T2 N0 M0 Success M AE fold SCC T2 N0 M0 Success M Palate SCC T2 N0 M0 Success M Tonsil SCC T1 N2 M0 Success M Tonsil SCC T1 N1 M0 Success M Tonsil SCC T2 N1 M0 Success F PPhyx Acystic T1 N0 M0 Success M PPhyx Acystic T1 N2 M0 Success M BOT SCC T2 N0 M0 Success M BOT SCC T2 N1 M0 Failure M False cord SCC T1N0 M0 Failure Abbreviations: SCC, squamous cell carcinoma; BOT, base of tongue; CR cell, clear cell carcinoma; PPhyx, paraparyngeal space; Acystic, adenoid cystic carcinoma; Outcome, defined by successful excision of tumor. 284 Transoral Robotic Surgery in Head and Neck Cancer HEAD & NECK DOI /hed March 2009

3 was inadequate and the procedure was terminated. The remaining 18 patients were clinically staged as follows: T1N0 (n 5 6), T1N1 (n 5 2), T1N2 (n 5 2), T2N0 (n 5 8). The primary sites included the tonsil (n 5 7), base of tongue (n 5 3), palate (n 5 2), posterior pharynx (n 5 1), supraglottis (n 5 3), and the parapharyngeal space (n 5 2). Histopathological diagnoses included squamous cell carcinoma (SCC) (n 5 15), adenoid cystic carcinoma (ACC) (n 5 2), and clear cell carcinoma (n 5 1). In 15 cases, the robotic cautery attachment was used in combination with the robotic scissors to achieve the resection. In 3 cases, the CO 2 flexible LASER was used in combination with the robotic bipolar forceps to achieve the resection. The LASER was fastened to robotic manipulator so that the tip could be manipulated The power setting of 10-W continuous mode was used for the procedures. The LASER proved ideal for cutting and coagulation of small vessels. It was also ideal for creating mucosal flaps as it allowed for fine mucosal and muscular incisions without significant peripheral tissue damage. In all 18 cases, frozen section was used to confirm negative resection margins. Margins were obtained from both the specimen (inked in the operating room at the time of surgery) and as separate specimens taken from the periphery of the resection bed. Negative margins were confirmed on formalin-fixed pathology in all 18 cases. Intraoral reconstruction was performed in 9 patients (Table 1). Nine patients underwent reconstruction. In 1 case following an extensive tonsillar resection, a mucosal advancement flap pharyngoplasty was performed. Two supraglottic resections were reconstructed with a pyriform mucosal flap. In 2 cases of palatal resections and 1 posterior pharyngeal wall resection, posterior pharyngeal wall flaps were performed for reconstruction. The remaining 2 pharyngeal resections required mucosal advancement flaps for primary reconstruction. In all cases, the robotic arms provided the dexterity to create mucosal and musculomucosal flaps to close the defects. In the 3 cases in which the CO 2 flexible LASER was used to create the flap, the mucosal flaps could be precisely fashioned facilitating the quality of the flap. The precision and control afforded by the CO 2 LASER was superior to that of the cautery and produced better hemostasis than incisions made with the scissors. In all cases, vicryl sutures (3.0, 4.0, and 5.0) were used for reconstruction. Sutures and knots were applied with the robot. Fifteen patients treated for SCC with TORS underwent concomitant unilateral (n 5 10) or bilateral (n 5 5) selective neck dissections immediately following the TORS procedure (Table 2). In 1 case, a communication between the pharynx and the neck required a musculo-mucosal flap for reconstruction of the defect. None of the patients required tracheotomy and there were no intraoperative or postoperative complications. Average estimated blood loss related to the robotic component of the surgery for the study group was 80 ml (range, ml). Table 2. Treatment and patient outcomes. Case Neck dissection Reconstruction Oral diet, day Discharge, day Radiation Chemotherapy 1 R Sel Pharyngoplasty 1 2 No No 2 L Sel None 1 2 No No 3 R Sel None 1 1 Yes Yes 4 R Sel None 1 1 Yes No 5 B Sel None 3 2 Yes No 6 B Sel Pharyngoplasty 4 2 Yes No 7 R Sel Palatoplasty 1 1 Yes No 8 None None 1 2 No No 9 B Sel None 1 2 Yes No 10 B Sel Mucosal flap 2 3 No No 11 B Sel Mucosal flap 1 3 No No 12 R Sel Palatoplasty 1 1 No No 13 R Sel None 1 2 Yes Yes 14 R Sel None 1 1 Yes No 15 R Sel None 1 1 Yes No 16 None Pharyngoplasty 1 2 No No 17 None Pharyngoplasty 1 1 Yes No 18 L Sel Pharyngoplasty 2 1 No Yes Mean Abbreviations: R Sel, right selective neck dissection; L Sel, left selective neck dissection; B Sel, bilateral selective neck dissection. Transoral Robotic Surgery in Head and Neck Cancer HEAD & NECK DOI /hed March

4 There were no cases of intraoperative or postoperative hemorrhages, and none of the patients required transfusion during or after surgery. Pathologically, 3 patients were upstaged. Pathological staging was T1N0 (n 5 4), T1N1 (n 5 4), T1N2 (n 5 2), T2N0 (n 5 6), T2N1 (n 5 1), T2N2 (n 5 1). All patients had pathologically negative margins as defined by no tumor or carcinoma in situ within the free margin, and there was no evidence of extracapsular spread in any neck dissection specimen. The decision to recommend adjuvant radiation in this cohort of patients was based on the presence of nodal disease, close margins (defined as less than 5 mm). Adjuvant chemotherapy was recommended for patients with extracapsular spread or tumors that exhibited perineural invasion, angioinvasion, or poor lymphoid response. Based on these criteria, postoperative radiation was administeredto10patients.thedosagewasreducedinall 10 patients from the standard therapeutic dose of 7200 cgy to an adjuvant dose of 6000 cgy (Table 2). Adjuvant chemotherapy was recommended for 4 patients, however only 3 elected to proceed with concomitant chemoradiation. Although follow-up is limited to an average of 5.1 months (range, 2 8 months), there has been no evidence of recurrence in any of the patients included in this study. Functionandqualityoflifedatafollowingadjuvanttherapy is not presented in this study. Preoperatively, all the patients included in this study were able to tolerate a regular diet. Postoperatively, all 18 patients began an oral diet 1.4 days average (range, 1 4 days) after surgery without clinical evidence of aspiration or velopharyngeal reflux (Table 2). All patients were evaluated by the speech therapy team preoperatively and postoperatively, prior to initiation of an oral diet. Endoscopic evaluation of swallowing to assess for aspiration, velopharyngeal reflux, and excessive laryngeal pooling of secretions was performed following surgery. Patients without evidence of aspiration were started on a puree diet on postoperative day 1 and advanced to a soft diet for 1 week on postoperative day 2. In our series, the 1 patient who underwent an extensive hypoharyngeal resection and 1 patient who underwent an extensive base of tongue resection were delayed in initiating oral diet until days 4 and 3, respectively. In all patients, voice quality and articulation were assessed 1 month following surgery as unchanged relative to preoperative assessment. The average length of hospital stay was 1.7 days (range, 1 3 days). All patients had clinical follow-up 1 month after surgery with FIGURE 2. The surgical setup time shown in chronological case order. Surgical setup time progressively decreased demonstrating the sharp learning curve. [Color figure can be viewed in the online issue, which is available at wiley.com.] barium swallow evaluation. None of the studies demonstrated aspiration or velopharyngeal reflux. Operating Room Time. Surgical setup time decreased dramatically as the operating room staff and physicians gained experience with the davinci Surgical System. The first case performed required 140 minutes to be set up. Over the course of the study, the setup time progressively decreased to 20 to 25 minutes setup time for most cases (Figure 2). In the study group, the average setup time was 54.6 minutes (range, minutes). Similar trends were observed in the analysis of the robot operating time. From TORS incision to closure, the mean surgical time was 84 minutes (range, minutes). DISCUSSION The application of robotic technology to surgery and medicine reveals a fascinating history. The term robot was coined by the Czech playwright Karl Capek in his 1921 play Rossom s Universal Robots. 11 Robot is the English translation of the Czech term robota, translated as forced labor. The term robot has subsequently evolved in meaning to suggest a form of artificial intelligence. Although robots have been used for several decades in industry, the crossover into medicine has been rather slow. Even though there are several possible reasons for this, it is becoming increasingly evident that robotic technology has a place in the field of surgery. Although the initial surgical robots were introduced to increase the precision of surgical procedures such as brain biopsies and fixation of long bone fractures 1,11 in the late 286 Transoral Robotic Surgery in Head and Neck Cancer HEAD & NECK DOI /hed March 2009

5 1980s, researchers at the National Air and Space Administration (NASA) Ames Research Center began to investigate the role of robotic telepresence surgery. Working with the Stanford Research Institute (SRI), scientists hoped to establish a program wherein surgeons could perform complex surgeries on wounded military soldiers from a remote location. Although this concept never came to fruition, the development of the telemanipulator, a component that allows the surgeon to manipulate the robotic arms using the motion of the wrist, thumb, and forefinger, led investigators to reconsider the potential application of robotic technology to surgery. Subsequently, robotic technology has been applied to a variety of surgical fields. Recently, investigators have evaluated the role of TORS for the management of head and neck tumors. Initial work has suggested that TORS is safe and potentially useful for the management of base tongue and supraglottic carcinoma. In our study, we found that initially, the positioning of the robot and placement of the robotic arms in a position that provide adequate range of motion was time consuming, however there is a sharp learning curve. After our initial cases, we found that the setup time precipitously decreased and leveled off at approximately 20 to 25 minutes. We found that with more experience, total operating room setup time, including sterile draping of the robot, anesthesia, positioning and draping of the patient, camera calibration, and introduction of the robotic arms in the oral cavity, decreased substantially. This was most evident from our first case which took over 140 minutes to setup, while the last several cases performed in the study group took just 20 minutes. O Malley et al 7 had similar results with total setup time ranging from 38 to 52 minutes. With 1 more year of experience, the same group published similar results with 3 patients undergoing supraglottic partial laryngectomy with setup timesrangingfromjust6to30minutes. 12 Similarly, actual operating time from incision to closure decreased dramatically with more experience, which correlates with other robotic surgical fields like urology and cardiac surgery. 13,14 Access to the tumor can also be challenging. Factors such as the anatomy of the patient, the type of retractor, and the site of the tumor impact on access. In our series of 20 patients, we were unable to achieve adequate surgical exposure in 2 cases. In our experience, the Crowe-Davis and Dingman retractors provide excellent access to the oral cavity, oropharynx, and the posterior pharyngeal wall; however, it does not provide adequate access to the larynx or hypopharynx. The FK retractor was effective in providing access to the supraglottic larynx in 2 cases; however the construct of the retractor precluded free unobstructed movement of the robotic arms in 1 case. In a second patient, we were not able to adequately gain access to a base of tongue cancer because of the patient s small, retrussive mandible. Currently, none of the retractors used in this study are ideal for exposure of the hypopharynx and larynx. In our series, we found that robot provides excellent visualization and manipulation of the tissue to facilitate surgical resection. The 3608 motion of the robotic instruments provided excellent access to cutting and suturing tissue a property that is not available in standard transoral LASER surgery. The most significant advantage of the robot was the ability to manipulate the tissue during the resection and suture the tissue during reconstruction. The ability to perform mucosal reconstruction eliminates the development of granulation tissue and provided control over healing, which we believe prevents velopharyngeal stenosis, nasopharyngeal insufficiency, and glotic stenosis. Weinstein et al 9 reported their experience with radical tonsillectomy and reported hypernasality in 1 patient that resolved following a transoral scar resection. The ability to reconstruct the oropharyngeal defect represents a significant advantage over other transoral techniques, and in our series, we did not experience measurable velopharyngeal insufficiency. Not dissimilar to the data reported with transoral LASER resections, 15,16 in our study, we found that patients tolerated the procedure with minimal morbidity, started an oral diet, on average, within 2 days of surgery and were discharged just 1.7 days following surgery. This data coincides with that reported by Weinstein et al. 9 It is interesting to note that age had no bearing on any of these parameters, suggesting that TORS is well tolerated in the aged population. TORS provides several advantages over open surgery and nonsurgical protocols. Similar to transoral LASER surgery, patients experience a more expedient recovery and do not require a tracheostomy. In an early study published by O Malley et al, 7 they performed tracheostomies as a safety precaution on their first 2 patients with tongue neoplasms; however, these patients were decannulated within 1 to 2 weeks after surgery. We did not perform a tracheostomy on any of the Transoral Robotic Surgery in Head and Neck Cancer HEAD & NECK DOI /hed March

6 patients in this study, and no patients exhibited airway compromise or required prolonged intubation all patients were extubated at the conclusion of the surgical procedure. The expedient recovery allows patients to begin adjuvant therapy, when indicated, earlier than in patients who undergo open surgery. This is important because as demonstrated by Suwinski et al, 17 who retrospectively reviewed 868 patients with head and neck cancer, on average, each day of extension between surgery and radiation was associated with a decrease in locoregional control by 0.17%. As cited by others, 15,16 1 advantage of complete resection of the primary tumor is that concomitant chemotherapy and high-dose radiation can be adjusted to treat patients with lower doses or external beam radiation and in select cases, withhold chemotherapy. Although organ-sparing protocols have been demonstrated to preserve vital organs such as the larynx and tongue, 80% of patients develop toxicity associated with therapy and a majority of patients suffer from both short-term and long-term functional morbidity. 18 Furthermore, numerous reports demonstrate that combined chemotherapy and radiation (doses of 7000 cgy and greater) used for primary therapy are synergistic with regard to toxicity 18 leading to unacceptably high rates of morbidity and mortality. 19 Despite the well-recognized morbidity associated with both nonsurgical and surgical approaches, several studies have demonstratedthatsurvivalisequivalentinbothtreatment protocols. A large retrospective analysis of 16,188 patients performed using the National Cancer Data Base demonstrated that survival was highest in patients who underwent surgery plus radiation when compared with either radiation alone or combined chemoradiation in advancedstage disease. 20 Such findings illustrate the need for a more effective, less toxic, and less functionally debilitating approach to therapy. In our study, we reserved radiation for patients with N2 or greater disease, and chemotherapy was concomitantly administered to patients with extracapsular spread or tumors exhibiting poor prognostic factors. This protocol was initiated in an attempt to reduce both short-term and long-term toxicity, without compromising survival. The data to determine the impact of this protocol on survival, function, and quality of life, are not yet available for interpretation. CONCLUSION In this pilot study, we have found that TORS is a safe, feasible, and minimally invasive alternative to classic open surgery or endoscopic transoral LASER surgery in patients with early cancer of the head and neck. With increasing experience, surgical setup time as well as operative time are manageable. Further prospective studies examining long-term functional outcome and disease control are necessary to determine if this technique provides a treatment advantage over currently available treatment regimens. REFERENCES 1. Kwoh YS, Hou J, Jonckheere EA, Hayati S. A robot with improved absolute positioning accuracy for CT guided stereotactic brain surgery. IEEE Trans Biomed Eng 1988;35: Davies B. A review of robotics in surgery. Proc Inst Mech Eng H 2000;214: Adili A. Robot-assisted orthopedic surgery. Semin Laparosc Surg 2004;11: Hollands CM, Dixey LN. Robotic-assisted esophagoesophagostomy. J Pediatr Surg 2002;37: ; discussion Boehm DH, Reichenspurner H, Gulbins H, et al. Early experience with robotic technology for coronary artery surgery. Ann Thorac Surg 1999; 68: Hockstein NG, Nolan JP, O Malley Jr BW, Woo YJ. Robotic microlaryngeal surgery: a technical feasibility study using the davinci surgical robot and an airway mannequin. Laryngoscope 2005;115: O Malley Jr BW, Weinstein GS, Snyder W, Hockstein NG. Transoral robotic surgery (TORS) for base of tongue neoplasms. Laryngoscope 2006;116: Weinstein GS, O Malley Jr BW, Hockstein NG. Transoral robotic surgery: supraglottic laryngectomy in a canine model. Laryngoscope 2005;115: Weinstein GS, O Malley Jr BW, Snyder W, Sherman E, Quon H. Transoral robotic surgery: radical tonsillectomy. Arch Otolaryngol Head Neck Surg 2007;133: Hockstein NG, O Malley Jr BW, Weinstein GS. Assessment of intraoperative safety in transoral robotic surgery. Laryngoscope 2006;116: Satava RM. Surgical robotics: the early chronicles: a personal historical perspective. Surg Laparosc Endosc Percutan Tech 2002;12: Weinstein GS, O Malley Jr BW, Snyder W, Hockstein NG. Transoral robotic surgery: supraglottic partial laryngectomy. Ann Otol Rhinol Laryngol 2007;116: Nifong LW, Chu VF, Bailey BM, et al. Robotic mitral valve repair: experience with the da Vinci system. Ann Thorac Surg 2003;75: ; discussion Pasticier G, Rietbergen JB, Guillonneau B, Fromont G, Menon M, Vallancien G. Robotically assisted laparoscopic radical prostatectomy: feasibility study in men. Eur Urol 2001;40: Steiner W, Fierek O, Ambrosch P, Hommerich CP, Kron M. Transoral laser microsurgery for squamous cell carcinoma of the base of the tongue. Arch Otolaryngol Head Neck Surg 2003;129: Grant DG, Salassa JR, Hinni ML, Pearson BW, Perry WC. Carcinoma of the tongue base treated by transoral laser microsurgery, Part 1: Untreated tumors, a prospective analysis of oncologic and functional outcomes. Laryngoscope 2006;116: Transoral Robotic Surgery in Head and Neck Cancer HEAD & NECK DOI /hed March 2009

7 17. Suwinski R, Sowa A, Rutkowski T, Wydmanski J, Tarnawski R, Maciejewski B. Time factor in postoperative radiotherapy: a multivariate locoregional control analysis in 868 patients. Int J Radiat Oncol Biol Phys 2003; 56: ForastiereAA,GoepfertH,MaorM,etal.Concurrentchemotherapy and radiotherapy for organ preservation in advanced laryngeal cancer. N Engl J Med 2003;349: Machtay M, Rosenthal DI, Hershock D, et al. Organ preservation therapy using induction plus concurrent chemoradiation for advanced resectable oropharyngeal carcinoma: a University of Pennsylvania phase II trial. J Clin Oncol 2002; 20: Zhen W, Karnell LH, Hoffman HT, Funk GF, Buatti JM, Menck HR. The National Cancer Data Base report on squamous cell carcinoma of the base of tongue. Head Neck 2004;26: Transoral Robotic Surgery in Head and Neck Cancer HEAD & NECK DOI /hed March