Transoral robotic surgery (TORS) in head and neck cancer: indications, limits and global technique 82
Transoral robotic surgery (TORS): selecting the basic instruments (other than the robot) for safe surgery 85
Transoral robotic surgery (TORS) for lateral oropharyngeal carcinoma 87
Transoral robotic surgery (TORS) for the management of tongue-base carcinoma 93
Transoral robotic surgery (TORS) for squamous cell carcinoma of the posterior pharyngeal wall 96
Transoral robotic surgery (TORS) for hypopharynx carcinoma 99
Transoral robotic surgery (TORS) for supraglottic laryngectomy 101
Transoral robotic surgery (TORS) and the vocal cords 103
Transoral robotic surgery (TORS) for total laryngectomy 106
Transoral robotic surgery (TORS) for benign tumors of the laryngopharynx 108
Transoral robotic surgery (TORS) for obstructive sleep apnea 111
Morbidity and mortality in transoral robotic surgery (TORS) 113
Transoral robotic surgery (TORS) has significantly improved the intraoperative and postoperative management of difficult-to-reach head and neck neoplasms. Prior to the development of TORS, open surgical approaches greatly impacted postoperative functional outcome and quality of life. These open procedures, when used to eradicate lesions of the oropharynx, hypopharynx, skull base and nasopharynx, involved long operative times, prolonged hospital stay and increased morbidity compared to TORS [ ].
In contrast to other minimally invasive surgical alternatives, such as endoscopic and microscopic techniques, robotic technology is utilitarian. Benefits associated with the da Vinci® Surgical System include 3D optics with high magnification and resolution, the assistance of more than one set of hands in the field, the ability to operate outside the direct line-of-sight, and a wide variety of operative instruments [ , ]. Its unique EndoWrist® technology affords specific advantages, including natural wrist freedom associated with 7 degrees of motion, absence of fulcrum effect, enhanced precision and tremor filtration [ ]. TORS may allow the operator to avoid external incisions, and it minimizes disruption of normal tissue planes compared to more invasive approaches [ , ]. Robotic systems can also be excellent teaching tools [ ]. Surgeons can mentor by sharing their view, providing two-dimensional telestration via freehand sketching over images, and proctoring via a dual console [ ]. The safety profile of this robotic application has been well validated in pre-clinical studies, animal subjects and human models [ , , , ]. TORS was approved by the Food and Drug Administration (FDA) in 2009 for otolaryngologic surgical resection of benign lesions, as well as malignant T1 and T2 head and neck tumors in adults [ ].
The major aim of TORS is to provide local disease control while optimizing functional outcome, quality of life and cosmesis. This includes decreasing the rate of postoperative tracheostomy and gastrostomy dependence [ ]. Another major goal is to provide a valid minimally invasive alternative to non-surgical therapies in order to minimize drug toxicity and associated morbidity, particularly in young patients [ , ]. When adjuvant treatment is necessary, dose reduction may be a reasonable goal in certain patients: e.g., with favorable pathologic risk stratification or a relatively small postoperative target tissue volume [ ].
In the early 2000s, robotic systems were used in ENT procedures in porcine and cadaveric models [ ]. McLeod et al. reported the first operation on a living patient, with excision of a benign vallecular cyst [ ]. Although McLeod et al. introduced the robotic arms via the mouth, the procedure was not a success because the authors used a laryngoscope and only two arms (the camera and one other instrument) to unroof a cyst. TORS, which for the first time utilized three robotic arms via mouth retractors, was developed at the University of Pennsylvania in 2004 [ ]. The team published results from the first human translational research subjects, who underwent robotic tongue base resection of T1 and T2 squamous cell carcinoma, in 2006 [ ]. The FK retractor was identified as the most versatile retractor for intraoral exposure, but the Crow Davis and Dingman retractors are valid alternatives. Orotracheal or nasotracheal intubation is performed, based on surgeon preference [ ]. Initially, 8 mm instruments were utilized, and the development of 5 mm instruments enhanced the versatility of the da Vinci® Surgical System. Typically, the surgeon operates at the console while an assistant is available at the head of the bed. If possible, the tumor is resected en bloc.
Current indications for TORS are primarily based on the absence of metastatic disease and favorable anatomic characteristics in patient and tumor in terms of accessibility, visibility and thus the probability of achieving negative margins. In 2009, the FDA cleared the da Vinci® system for T1 and T2 oral, pharyngeal and laryngeal cancers and benign tumors. It is not currently FDA cleared for T3 and T4 lesions [ ]. However, TORS has been employed for appropriate candidates with advanced stage disease including up to T4a oropharyngeal lesions [ , ]. In the oropharynx, TORS is contraindicated if neck disease is not resectable, if there is evidence of mandibular invasion or carotid artery or prevertebral fascia involvement, or if the tumor involves more than 50% of the tongue base or posterior pharyngeal wall [ ]. In the larynx, surgical limitations also include accessibility and size [ ]. Resectability of laryngeal lesions, in particular, must be balanced with functionality compared to non-surgical alternatives [ ]. Approaches and indications for skull base tumor resection are being developed in a stand-alone fashion, as well as in combination with endoscopic techniques [ ].
TORS for obstructive sleep apnea (OSA)
In 2010, the role of TORS was expanded by Vicini et al. to include base of tongue resection for the treatment of OSA [ ]. TORS for OSA has since progressed to include resection at multiple levels, including lingual tonsillectomy, uvulopalatopharyngoplasty (UPPP), supraglottoplasty, Z-palatoplasty (ZPP), uvulopalatal flap, limited pharyngectomy, and epiglottectomy or epiglottoplasty [ ]. TORS provides easy exposure of the hypopharynx, and this consistent ease of exposure facilitates surgery for OSA. TORS posterior glossectomy and limited lateral pharyngectomy with UPPP have been shown to significantly decrease apnea duration and reduce apnea-hypopnea index (AHI) by more than 50% [ ]. Midline glossectomy and ZPP were more effective than other minimally invasive alternatives (radiofrequency treatment or submucosal coblation) in reducing AHI [ ]. As multilevel surgical approaches to manage airway obstruction in sleep apnea continue to evolve, TORS promises marked improvement in treatment outcome compared to continuous positive airway pressure (CPAP), the current gold-standard therapy, for a subset of suitable patients [ ].
Exciting technological advances are on the horizon for TORS, many of which revolve around improvement in visualizing and accessing the operative site. The da Vinci® SP™ Surgical System introduces the possibility of a single port approach, as well as flexible endoscopes [ ]. Its initial FDA clearance is currently for urologic procedures, but it may ultimately present a new landscape for the evolution of TORS. Moore et al. described another innovative method to improve visualization in TORS, using currently available technology and placing the patient in a seated position, thereby improving line-of-sight access to the inferior pharynx and larynx [ ].
The future of TORS will lead to new devices and operative frontiers. Teletransmission of surgical movement through a robotic system may enable surgical expertise to be available in remote locations and improve surgical training worldwide [ ]. Though shown to be feasible, telesurgery is not widely implemented, due to practical and legal restrictions. Nevertheless, collaboration and innovation will continue to enable TORS to impact future patient care. Its limits will be defined only by the imagination of its patrons.
Transoral robotic surgery (TORS) in head and neck cancer: indications, limits and global technique *
* Bernier J (Ed). Head and Neck Cancer. Multimodality Management. Springer ; 2016. p. 388–390. Reproduced with permission of Springer.
The development of partial laryngeal surgery in the 1970s [ ] and endoscopic laser surgery in the 1990s [ ] aimed to reduce functional and esthetic sequelae while maintaining oncological efficacy. The development of transoral robotic surgery (TORS), which emerged in 2005 [ ], was also initiated in an attempt to achieve this goal. B. O’Malley and G. Weinstein’s team in Philadelphia (Pennsylvania, USA) used the da Vinci® robot to treat a benign oropharyngeal tumor after a feasibility study was conducted on a porcine model [ ]. Studies assessing TORS for the treatment of upper aerodigestive tract cancer were published as early as 2008 [ ]. Current indications for robotic surgical procedures include T1-T2 tumors of the oropharynx and the supraglottic larynx.
However, this type of endoscopic surgery involves difficulties associated with the robot-related materials (robot arms and instrumentation), patient apprehension, tumor exposure and the individual surgeon’s learning curve. The aim of this text is to update the current status of robotic surgery for the treatment of cancers of the upper aerodigestive tract.
The principles of TORS
All TORS procedures are performed under general anesthesia and oral or nasotracheal intubation. The da Vinci® robot has several components: the robot itself, a surgeon’s console, light sources, 3D camera and mono- and bipolar generators. It is recommended to use one of the three specific mouth gags available (FK Olympus®, LARS retractor Fentex® and M Micro France®). These provide large access to the pharynx in order to place the 8 mm optic at an angle of 0° or 30° inside the pharynx, using the two robotic arms equipped with instruments (Maryland forceps and monopolar electrocautery) ( Figure 7.1 ). The arms are controlled by the surgeon, who works near the patient from a console with 3D vision and up to 10-fold magnification. Installation time (approximately 20 to 30 minutes) for the retractor and arms is essential. Good exposure of the tumor is the key to a successful TORS procedure. An assistant is always positioned at the patient’s head to aspirate fluids and fumes and monitor the position of the arms inside the mouth so as to prevent potential conflict. An operating room nurse loads and cleans the instruments on the robot arm. At the console, the surgeon manipulates two joysticks that transmit motion to instruments with a ratio of 5 to 1. This enables high-precision movements and eliminates tremor. Using the joysticks is very natural, allowing the operator to work in a comfortable sitting position. It is necessary to reposition the retractor and robot arms during tumor resection. Resection uses the monopolar forceps, and hemostasis uses the Maryland forceps. The wound is left to control healing or sutured with a local flap, depending on its location.
As stated by Weinstein et al. [ ], preoperative endoscopy can be used to select patients eligible for robotic surgery. The surgeon who will perform the robotic surgery using the dedicated retractor must also perform the endoscopy. The main criteria assessed on preoperative endoscopy are the anatomical location and size of the tumor, mouth opening and maxillo-mandibular anatomy. Small mouth opening and/or large tongue base are contraindications that only the surgeon can evaluate. With this pre-selection, a very small number of patients prove ineligible due to poor exposure. Only 13 out of 129 patients had inadequate exposure in a multicenter study by Weinstein et al. [ ].
The Food and Drug Administration (FDA) approved TORS only in T1-T2 tumors of the oropharynx, larynx and hypopharynx. The absence of tactile force feedback makes palpation impossible with the robotic instruments, which makes highly invasive tumors ineligible, in our experience. The size of the robotic instruments (8 mm) is too large for working in the endolarynx, where endoscopic laser surgery is preferred. Selection of patients eligible for TORS must be approved by the multidisciplinary team meeting (oncologist, radiologist, histologist and head and neck surgeon) on the basis of the results of the endoscopic evaluation, cervical and thoracic CT scan and histologic diagnosis.
Start-up in robotic surgery cannot be improvised and must be planned so that the whole team is properly informed before initiating the first case. Health institutions that invest in an expensive robotic system should routinely offer this training. Ideally, the first case should be scheduled within 2 weeks of training. In addition, the first case should be easy (small tumor of the tonsil) and supervised by an experienced TORS surgeon. Under these conditions, the learning curve is quite fast. Several publications [ , ] have shown that TORS duration for a given surgeon decreased significantly after the first 10 patients. The surgical team must follow a progression in the programming of cases, beginning with lesions of the oropharynx, followed by the supraglottic larynx, and finally the hypopharynx.
Anesthesia and operative specificities
The anesthesia required for TORS is not different from that required for other endoscopic surgeries. We prefer nasal intubation, to avoid placing the probe in the oral cavity. Weinstein et al. performed oral intubation, placing the probe in the contralateral labial commissure [ ].
The quality of resection margins is an important prognostic factor in local control. The use of frozen section is highly recommended when there are short margins.
In a TORS series from a group of surgeons in France [ ], neck dissection was performed in the same step in most cases but, for larger tumors with risk of cervical fistula, it can be postponed for 2 weeks. Where there is significant risk of intraoperative bleeding (risk of injury to the lingual artery), neck dissection can be performed first, to control and bind the branches of the external carotid that risk being injured.
The large number of publications on TORS confirms the growing role of this new technology for the surgical treatment of head and neck cancer. Team training, assessment of tumor exposure and adherence to oncological rules are essential criteria for this less invasive surgery. The contribution of technical innovations is expected to partially compensate for the absence of tactile force feedback. The use of reconstruction surgery by local or free flaps is also under development. Finally, the future of robotic surgery in our discipline lies in the development of new systems dedicated to TORS, with articulated thinner arms and flexible endoscopes. Several projects are underway and are expected to become available in the coming years.
Advantages of TORS
A 3D endoscope at 0° or 30° provides large access to the pharyngeal cavity.
TORS enables high-precision movements and eliminates tremor.
TORS for T1 and T2 pharyngeal tumors has the same oncological results as conventional open surgery.
Limitations of TORS
TORS needs team training and assessment.
Preoperative endoscopy must be performed by the TORS surgeon to select patients eligible for robotic surgery.
The use of a dedicated mouth gag is recommended.
The absence of tactile force feedback makes palpation impossible with the robotic instruments; highly invasive tumors are therefore ineligible.
Transoral robotic surgery (TORS): selecting the basic instruments (other than the robot) for safe surgery
Robot-assisted head and neck surgery has developed over the past 10 years, with constant adaptation not only of the specific instrumentation (the robot itself) but also of the non-specific instrumentation used in associated procedures, whether scheduled or emergency.
There are currently two robotic surgery systems available on the international market. The older, which has been available for a decade, is the da Vinci® robot in its successive versions. The more recent model, available since 2014, is the Flex® robotic system, designed exclusively for use via the orifice.
Transoral procedures require an operative assistant at the patient’s head ( Figure 7.2 ) to ensure suction of smoke and blood and take hold of the resection specimen with instruments adapted to the presence of the robotic instruments in the operative field.
We will consider successively the instruments used for introducing the robotic arms (retractors), suspension instruments, and those for additional procedures such as lymph node dissection, surgical reconstruction or for emergency cases (hemostasis, airway release).
According to the situation and surgeon’s preferences, minimalist or invasive retractors may be used. Retractors have been well described by O’Malley et al. [ ]. They range from the Boyle Davis ( Figure 7.3 ), Dingman or FK retractor to specific retractors that are being constantly improved, such as S. Morinière’s Retractor ( Figure 7.4 ) or the newly designed retractor for the Flex® robotic system.
Suspension systems have progressed from the single shelf to specific systems (Flex)®. Stabilizers used in digestive surgery (e.g. Omnitract Integra®) ( Figure 7.5 ) can also provide complete stabilization for transoral or transcutaneous retractors.
Transoral or transcutaneous surgery is performed in a confined space where smoke and mist on the endoscope can hinder the surgeon’s movement. Continuous suction is therefore necessary, but must not interfere with the procedure (instruments, sight). Suction can be directly included in the retractor ( Figure 7.6 ) or use a long flexible suction cannula ( Figure 7.7 ).
Equipment for additional surgery (lymph node dissection, tracheostomy, hemostasis)
Lymph node dissection is often associated to transoral tumor surgery and requires basic head and neck surgery instruments, and hemostasis instruments such as a mono- or bipolar electrocautery or hemoclip. It is also recommended to have instruments for emergency tracheostomy or cervical vascular surgery in the operating room in case of uncontrolled transoral bleeding.
It is necessary to have a good arm support to put the patient’s arm in abduction [ ]. Gynecologic or autostatic retractors can be used to help the manipulation of the robot instruments.
An operating table synchronized with the robot needs to be developed for head and neck robotic surgery, as used in abdominal robotic surgery. Currently, the operating table should allow secure rotation and roll to position the robot’s base and arms according to the patient’s morphology.
Robotic equipment needs an operating room providing adequate space for the robot. The operating room is used by several surgeons, with discussion before beginning robotic surgery. Various requirements have to be discussed (area, floor, connectors, etc.).
There is no specific anesthesia equipment, but transoral robotic surgery requires the anesthesia team to be at the patient’s foot because of the space taken up by the robot.
Transoral robotic surgery (TORS) for lateral oropharyngeal carcinoma
Introduction and history of the treatment of oropharyngeal tumor
The incidence of upper aerodigestive tract tumor, taking all locations together, seems to be slightly decreasing overall, but that of oropharyngeal tumor in particular seems rather to be slightly on the increase. This is in line with the particular epidemiological pattern of reduced alcohol consumption and smoking but an increase in HPV-related carcinoma [ ].
Historically, oropharyngeal cancer was systematically treated by often heavy surgical resection (cervico-mandibulotomy), particularly before the 1940s. Radiation therapy and then chemoradiotherapy gradually emerged as effective options, in direct competition with surgical indications. Indications for chemoradiotherapy, developed in the 2000s, gradually extended as oncology assessment demonstrated results equivalent to those of surgery, with reduced morbidity.
In a review of the literature in 2002, Parsons et al. reported that oncologic results in operable oropharyngeal tumor were equivalent between surgery and concomitant chemoradiotherapy [ ].. Thus, between the 1990s and 2010, non-surgical treatment of oropharyngeal tumor has increased as rates of surgical resection decreased ( Figure 7.8 ) [ ].
American results were corroborated by the most recent update of the 94-01 GORTEC study [ ] in Europe, confirming the advantage of non-operative treatment associating concomitant platinum-based chemotherapy and radiation therapy.
It was basically because of the morbidity associated with surgery that non-operative management became more widespread for the treatment of locally operable oropharyngeal tumor since the 1990s. The rate of early complications in classic open surgery (i.e. oropharyngectomy via mandibulotomy) before the advent of TORS was estimated at nearly 25%, compared to less than 10% with concomitant chemoradiotherapy. Indications for oropharyngectomy have progressively shifted from first-line attitude to salvage [ ].
The oncologic benefit of associating chemotherapy to radiation therapy in oropharyngeal cancer is now established, especially in locally advanced forms [ ]. Nevertheless, the side-effects of non-operative treatment are well known: rates of grade 3 or 4 mucositis rise from <40% to > 70% when chemotherapy is associated concomitantly to radiation therapy [ ], requiring gastrostomy and support care for more than 25% of patients [ ].
TORS has progressively developed as first-line treatment for oropharyngeal tumor since it received FDA clearance in 2009. Like with laser surgery some years earlier, the interest lay in limiting surgical morbidity. All the reports published since the first TORS procedures indicate the feasibility of therapeutic de-escalation, although there have as yet been no randomized studies comparing TORS versus concomitant chemoradiotherapy.
Since 2009, TORS indications have significantly increased, by 7.9% and 11.3% respectively, for AJCC (American Joint Committee on Cancer) grades I and II tumor [ ] while indications in grades III and IV have remained stable ( Figure 7.9 ). The same pattern holds comparing stage T1/T2 versus T3/T4.
The current trend is thus a return to resection, but by TORS, when this approach is feasible and resection can be deemed functional. TORS is especially favored for lesions otherwise inaccessible to surgery without mandibulotomy (T2 electively). In this situation, the morbidity of open surgery contrasts with non-operative concomitant chemoradiotherapy.
TORS: feasibility and indications
The main indication for TORS is resection of oropharyngeal tumor. In the 2012 multicenter study of TORS by Weinstein et al., 78% of the tumors were oropharyngeal, and 95% malignant [ ]. The French Group of Transoral Robotic Surgery [ ] reported 169 patients treated between 2009 and 2014; tumor stages were T1 in 51 cases, T2 in 100, T3 in 16 and T4 in 2; 6.7% of patients could not be managed by TORS, due to exposure problems. The authors all consider that these can be anticipated on preoperative endoscopy performed by an operator with experience in TORS, for prior assessment of the feasibility of positioning a retractor, reducing the rate of such intraoperative problems to less than 1% [ ].
TORS and resection margins
In Kelly et al.’s literature review [ ] of 11 relevant publications, all studies used systematic frozen section biopsy. Five of the 173 oropharyngeal specimens (2.9%) were intralesional on definitive histology (T1/T2 tumors); the patients underwent surgical revision, achieving final negative margins.
TORS and lymph node treatment
Kelly et al., in their literature review [ ], examined the treatment of lymph-node areas in T1/T2 oropharyngeal tumor managed by TORS. Most authors recommended lymph-node treatment in the same step as TORS, while a minority deferred it to limit the risk of mucosal breach [ ]. There no longer seems to be a good argument for dissociating TORS and neck dissection; indeed, one advantage of the one-step attitude is the possibility of access to the lingual vascular pedicles for intraoperative ligation if necessary [ ].
TORS, tracheotomy and enteral feeding
Attitudes toward systematic tracheotomy
In seeking to reduce morbidity, it has been natural to reduce the rate of intraoperative tracheotomy during TORS. In the 2012 North American study based on results from the MD Anderson Cancer Center, the Mayo Clinic and the Universities of Pennsylvania and Alabama, incidence of intraoperative tracheotomy was low, but highly variable from team to team [ ], ranging from 5% to nearly 30% for planned tracheotomy; rates of definitive tracheotomy, on the other hand, were consistently less than 2%. Yeh et al., in a review of the literature comparing TORS versus intensity-modulated radiation therapy (IMRT) (in non-randomized studies), found no significant difference in long-term tracheotomy according to treatment, with rates ranging between 0 and 4.5% in both [ ].
Attitudes toward feeding
Iseli [ ] reported 68% of patients with exclusively oral feeding at discharge and 86% by day 14, with significantly higher rates in treatment of second primaries or recurrence. In Yeh et al.’s literature review [ ], prolonged enteral feeding rates ranged between 0 and 20.7%. These rates are just slightly poorer than those for non-operative treatment associating IMRT and chemotherapy. Comparison is hindered by the lack of randomized studies. According to Kelly, the rate of dependence on enteral tube feeding, even partial, beyond 1 year was less than 5% [ ].
TORS and radiation therapy/chemotherapy
Adjuvant radiation therapy/chemotherapy
Although there is no consensus specific to TORS, most authors indicate adjuvant radiation therapy on classic criteria founded on postoperative pathology examination of the resection specimen (perineural invasion, vascular emboli, margin quality) and neck dissection findings (presence and number of invaded adenopathies, capsule rupture). In Weinstein et al.’s 2012 series of T1/T2 oropharyngeal tumor treated by robotic surgery and neck dissection without adjuvant radiation therapy, only 3% of patients showed locoregional recurrence [ ]; more than 80% of these were squamous cell carcinoma patients with negative margins (in some cases following revision surgery). In Smith et al.’s series[ ] in which 90% of the tumors were stage T1-T2, 21% of patients received postoperative adjuvant radiation therapy and 31% concomitant chemoradiotherapy; adjuvant treatment did not significantly improve survival. Iseli [ ] reported that 41% of patients managed by TORS received postoperative adjuvant radiation therapy.
Preoperative radiation therapy
In Iseli’s series [ ], 22% of patients undergoing TORS had had prior radiation therapy; Aubry et al. [ ] reported a rate of 31%, without evidence of increased postoperative risk.
TORS: role and results
Yeh et al.’s literature review [ ] compared results in oropharyngeal cancer treated by surgery versus isolated IMRT in more than 40 studies; these were neither randomized nor controlled and tumor size tended to be smaller in the TORS groups; nevertheless, the data pointed to equivalence in oncologic results ( Table 7.1 ) and slightly better functional results with TORS.
|Author||N||Included stages||Median follow-up (years)||Overall survival (%)|
|Broglie et al.||124||T1-T4||3||69|
|Chao et al.||26||Any T, any N||2||100|
|Daly et al.||107||Any T, any N||3||83|
|Eisbruch et al.||67||T1-T2, N0-N1||2||96|
|Garden et al.||776||Any T, any N||5||84|
|Huang et al.||71||Stage 3/4||3||83|
|Ingle et al.||24||Stage 3/4||2||92|
|Nichols et al.||44||Stage 3/4||3||79|
|May et al.||170||Any T, any N||3 *||87|
|Mendenhall et al.||130||Any T, any N||5||76|
|Saba et al.||65||Stage 3/4||3||83|
|Setton et al.||442||Any T, any N||3||85|
|Shoushtari et al.||112||Any T, any N||3||77|
|Yao et al.||66||Any T, any N||3||78|
In Kelly et al.’s meta-analysis [ ] of more than 200 articles comparing TORS versus concomitant radiation + chemotherapy in T1/T2 oropharyngeal tumor, there was no significant difference in survival according to treatment. Local control rates approximated 90% in both cases ( Figure 7.10 ). In Smith’s series, 3-year overall survival was 94% [ ].
Taken together, these studies tend to show oncologic equivalence between TORS and non-operative treatment. Functional comparison is greatly hindered by the lack of randomized studies. TORS, however, may, for an equivalent quality of results, avoid irradiating selected patients without indications for adjuvant radiation therapy, especially in small tumors that can be resected completely, without lymph-node invasion or histologic factors for poor prognosis.
In Weinstein et al.’s 2012 North American multicenter study, 95% of patients recovered full oral feeding [ ]. Iseli’s figures [ ] were slightly lower, at just over 80%. Definitive swallowing disorder, assessed on functional scores, affected a minority of TORS patients; poor scores correlated with tumor size, treatment context at recurrence, and preoperative swallowing disorder. Differences between series related to analysis methodology and whether food was strictly normal or adapted.
TORS: risk assessment and prevention of complications
Complications associated with TORS have been the focus of publications by several teams reporting their experience. During the first years, there were many small series, and this early information was very variable. It appeared that intraoperative complications were few, but that rates during the first month could be as high as 40% [ , , , ].
According to Mercante et al. [ ], intraoperative complications are exceptional, and not specific to TORS: anaphylactic shock, anesthesia-related problems, etc. The literature data on complications, however, are seriously dependent on the severity thresholds employed. Aubry et al. [ ] reported a 6.7% rate of intraoperative complications, about half of which consisted of bleeding, without large vessel lesion, amenable to simple endoscopic hemostasis; hemorrhage risk correlated significantly with anticoagulation/antiplatelet therapy; age and prior radiation therapy were not risk factors. Vergez et al. reported rates exceeding 11% [ ].
Rates of crossover to cervicotomy are consistently low, ranging between 0 and 3% [ , ], but higher in case of intraoperative opening of the pharyngeal mucosa, notably in hypopharyngeal tumors.
In 2016, the American College of Surgeons’ National Surgical Quality Improvement Program published an analysis of more than 300 patients treated by TORS for oropharyngeal cancer in the USA between 2010 and 2013 [ ]. Thirty-day mortality was low, at 0.7% in the group’s experience. Risk factors comprised elevated blood pressure, dyspnea and pre-anesthesia ASA score of 3 or 4, and correlated significantly with hospital stay exceeding 4 days. Risk was elevated in tongue-base locations compared to the lateral oropharynx, but those patients were also significantly older, with poorer ASA scores than patients with tumor in the tonsillar region; The French group reported post-TORS mortality of 1.1%.
In comparison (although studies were non-randomized), 1-month mortality after exclusive radiation therapy for equivalent tumors seems similar, at 2-3%, depending on the fractioning regimen [ ].
Postoperative course and complications
Postoperative complications rates range between 5% and 10%. Aubry and Smith both reported rates of 7% [ , ]. Su et al., reporting several hundred cases, found about 8% [ ].
Postoperative bleeding is a known risk, accounting for 7-8% of post-TORS complications [ , ]. Unaccountably, Aubry et al. [ ] found a rate of 18%, but in a series including many extra-oropharyngeal locations and patients under anticoagulation/antiplatelet therapy.
Su et al. [ ] reported a 1.6% rate of postoperative bleeding requiring transfusion, and higher in tongue-base locations. The authors stressed that transfusion rates tend to be underestimated, being counted as complications only in case of bleeding requiring surgical revision; they discuss this issue exhaustively, with 2 deaths at least partly implicating postoperative bleeding. In Weinstein et al.’s 2012 multicenter study of almost 200 patients, on the other hand, no requirements for transfusion were reported.
Pneumopathy is one of the most frequent complications, at 3%, one-third requiring intubation for assisted ventilation, according to Su et al. [ ]. Rates are higher in series including all locations, notably larynx and hypopharynx [ ].
Secondary emergency tracheotomy rates are low, at 5%, mainly concerning supraglottic resection [ ]. Postoperative high-dose corticosteroids are sometimes used to reduce the risk of tracheotomy, although there is no consensus. Tracheotomy is one of the most frequently highlighted forms of morbidity, as it may be avoided by use of TORS. It is nevertheless a means of preventing hemorrhage and severe edema with risk of asphyxia, and should be considered by the surgeon according to elements in the pre-treatment assessment: pre-existing edema, hemorrhage risk, laryngeal tumor location, etc.
Onset of cervical spondylodiscitis is very rare, mainly implicating resection involving the prevertebral fascia of the posterior pharyngeal wall or radiation therapy. In such extended resection, some authors recommend biological glue, or local cover flaps [ ].
TORS treatment of stage T1/T2 oropharyngeal tumor has become standard practice, although there have been no comparative studies against non-operative treatment. Large-cohort literature reviews suggest that oncologic and functional results are close to those of non-operative treatment. The benefit over resection via cervicotomy in terms of morbidity and mortality, on the other hand, is considerable.
One-step association of endoscopic tumor resection and neck dissection is recommended by most teams, and seems consensual.
A learning curve is necessary [ ], as is suitable training in the technique.
Many authors report the feasibility of isolated TORS without complementary therapy in T1/T2 tumor with complete resection on pathology, without massive lymph-node invasion. In that case, the advantage of TORS is to avoid radiation therapy without loss of benefit.
Postoperative risks are non-negligible: inhalation pneumopathy, secondary hemorrhage, or inspirational dyspnea. Being aware of these and of the risk factors (tumor size and location, anticoagulation therapy, type of pre-treatment feeding, etc.) guides possible indications for temporary tracheotomy and/or enteral tube feeding.
Transoral robotic surgery (TORS) for the management of tongue-base carcinoma
Due to the morbidity associated with open approaches to the tongue base, and because transoral robotic surgery (TORS) provides a binocular 3D HD view with improved visualization of anatomical structures, many surgeons find that tongue-base surgery is the most useful application of TORS. The base of tongue and lateral wall of oropharynx are the most common indications for TORS, and many studies have demonstrated its efficacy here [ ].
Techniques, feasibility, quality
A multicenter study by Weinstein et al. established the safety and feasibility of TORS [ ].
The use of a mouth prop is the main commonly reported technique, but some have described suture through the mobile tongue with distraction and suspension of the tongue onto the operating table, to avoid needing any prop [ ].
Several studies described transoral visualization of tongue-base anatomy, in anatomic studies of cadaveric specimens with an endoscopic approach, to improve knowledge of specific landmarks and safety [ ].
Since O’Malley reported the first three cases of tongue-base TORS resection for carcinoma in 2006 [ ], few authors have focused on tongue-base carcinomas, with most studies concerning oropharyngeal lateral wall resection or supraglottic laryngectomy. TORS is FDA approved for T1 and T2 tumors. In the series of the NCDB, patients male sex and lower T stage were mainly associated with a TORS primary treatment [ ].
Most reported studies were case series with small samples. They demonstrate feasibility, with reasonable average procedure times, acceptable postoperative complication rates and good functional outcomes mainly for T1–T2 tumors [ ]. The N stage does not modify the resecability of small tumors but impact the probability of adjuvant concurrent chemoradiation in case of poor pathological features. In this way high N stage are rarely treated with TORS.
Carcinoma from unknown primary
Recently TORS was employed to locate the primary tumor in occult primary carcinoma with specific focus on the tongue-base mucosa [ ]. Given the rise in HPV-related oropharyngeal cancer, the percentage of patients with unknown primary squamous cell carcinoma (SCC) is increasing. According to Motz et al., the incidence of unknown primary SCC is increasing; the majority are HPV-related, and may be localized to the tonsils and base of the tongue [ ]. The goal is to either treat the unknown primary surgically, for small oropharyngeal tumors, or to limit radiation dose to the tumor site.
In an analysis of 22 carcinomas from unknown primary (CUP), Byrd et al. reported an 86.4% rate of location of the primary by robot-assisted HD examination of the oropharyngeal mucosa [ ]. Other studies reported similar success [ , ]. Systematic removal of the lingual tonsil has been recommended [ ].
Recently, Fu et al., in a systematic review of the literature, identified 139 patients with CUP managed by a transoral approach. The primary could be identified in 67% of patients, with no remarkable findings on physical examination, radiologic imaging or panendoscopy with directed biopsies. Moreover, lingual tonsillectomy identified the primary tumor in 18/25 (72%) of patients with no findings [ ]. Thus systematic lingual tonsillectomy must be considered in patients with CUP.
De Almeida et al., in a systematic review comparing TORS-based treatment to non-surgical treatment, reported that 2-year oncologic outcome did not differ but TORS led to better functional outcome [ ]. More recently, a meta-analysis found that 5-year overall survival and disease-specific survival were equivalent between the two strategies [ ]. In a multicenter study of 410 patients De Almeida et al. reported 91.8% 2-year locoregional control and 94.5% disease-specific survival; only 21.3% of patients received adjuvant chemoradiation [ ].
In a meta-analysis, Morisod and Simon found 89.6% 5-year overall-survival in 276 patients treated by transoral surgery [ ]. Garden et al. reported that base of tongue primaries had improved overall survival compared to primaries from the lateral pharynx and soft palate regardless to the treatment [ ].
Several well known clinical or pathological features may increase the rate of positive margins. Only one study has focused on the specificity of TORS. In a study of 3,071 patients from the US National Cancer Database, Cracchiolo et al. reported that positive margin rates were lower when TORS was performed in a high-volume rather than low-volume hospital (8.2% versus 16.7% respectively; p=0.001). In this series, 14.2% of patients had a positive margin resection when treated with TORS compared to an overall positive margin rate of 23.8% in patients treated with other surgical approach [ ]. A systematic review protocol on oncological and survival outcomes following TORS versus transoral laser microsurgery is currently ongoing [ ].
To evaluate quality of life (QoL), swallowing and voice specifically in patients with base of tongue (BOT) carcinoma, Mercante et al. reported a prospective cohort trial of 13 patients. Dysphagia score (DS), fiberoptic endoscopic swallowing assessment on the penetration aspiration scale (PAS), MDADI (MD Anderson Dysphagia Inventory) and VHI-10 (Voice Handicap Score-10) score were assessed at baseline and 6 and 12 months. Objective swallowing deteriorated during the first 6 months after TORS alone, but with complete recovery within 12 months and normal QoL scores at 1 year [ ].
In a systematic review analyzing 441 patients, Hutcheson et al. reported chronic gastrostomy-dependence rates of 0% to 7%. MDADI scores ranged from 65.2 to 78, and incidence of postoperative pneumonia from 0% to 7%.
Temporary tracheostomy is systematic for some authors [ ] while others never perform it [ ]. In Weinstein’s study, only 26% of the patients underwent tracheostomy [ ]. Mockelmann, in a study of 42 patients, suggested a two-stage procedure, with delayed neck dissection to decrease the risk of tracheostomy; patients remained intubated for one night in the intensive care unit (ICU) following TORS, except those managed with elective temporary tracheotomy, which was only performed in 3 high-risk patients [ ].
TORS is associated with shorter hospital stay, less gastrostomy tube and tracheotomy placement, and lower hospital-related costs than open surgery [ ].
Timing of neck dissection in TORS
In a series of 113 patients, Kucur et al. reported the safety and efficacy of neck dissection concurrent to primary TORS, and found benefit in terms of avoiding pharyngocutaneous fistula compared to open surgery [ ]. However, some recent reports used a two-stage procedure with delayed neck dissection in order to decrease the risk of pharyngocutaneous fistula and tracheostomy. In a 41-case prospective series, Mockelmann et al. compared concurrent tumor resection and neck dissection to delayed neck dissection, 8.4 days after TORS. Complications rates and oncological quality indicators of neck dissection were similar in the two groups [ ].
TORS versus chemoradiotherapy
Open surgical approaches to the oropharynx can be associated with morbidity such as cosmetic deformity, malocclusion and dysphagia. Therefore, a trend for primary radiotherapy and concurrent chemoradiotherapy (CCRT) has been observed in the last few decades. However, to date no clinical trials compared surgery-based to radiotherapy-based treatment. Moreover, the toxicity of intensive chemoradiotherapy has been reported to lead to severe dysphagia, long-term gastrostomy-dependence, quality of life impairment and reduced eligibility for salvage treatment [ ]. The main goal of transoral surgery is to spare patients the long-term side-effects of CCRT, which have not been extensively studied in survivors more than 2 years after treatment. This is of particular concern for patients with HPV-positive oropharyngeal cancer, some of whom may be expected to live 20 to 30 years after treatment [ ].
When compared to IMRT, survival outcomes are comparable and functional outcomes may be superior but rationale is poor and only based on uncontrolled reports. Benefits of TORS may be largely due to the ability to deintensify adjuvant treatment [ ]. The earlier age of presentation and improved survival observed in HPV-related OPSCC has highlighted the morbidity risks associated with adjuvant treatment and prompted clinical trials examining the role of transoral approaches as part of a de-escalation strategy [ ]. In a meta-analysis comparing 729 patients treated by radiotherapy and 276 treated by a transoral approach Morisod et al. found similar 5-year disease-specific and overall survival rates [ ]. The question of adjuvant therapy in patient with more advanced N stage is jeopardized by retrospective study demonstrating that extracapsular extension does not carry the same prognosis for oropharyngeal cancers [ ]. This underline the need for prospective studies in specific field of oropharyngeal carcinomas with stratification based on HPV status and primary treatment.
TORS in salvage strategy
In patients who require salvage TORS after chemoradiation, additional considerations include airway edema, postoperative hemorrhage, potential fistula, velopharyngeal insufficiency, and carotid exposure [ ].
However, successful salvage TORS has been reported, with low complication rates without routine reconstruction [ ]. Local flap and free tissue transfer can be used to minimize the above risks [ , ].
Perspectives (on-going trials)
To date, no prospective trials have compared TORS with adjuvant therapy versus (chemo)radiation, although several are currently underway [ , ].
The RTOG 1221 trial was prematurely closed owing to lack of inclusion. This trial was designed for early T-stage, p16-negative OPSCC (orophayngeal squamous cell carcinoma). However, patients presented with more advanced disease, unresectable on a transoral approach [ ].
The ECOG 3311 trial is an on-going phase-II randomized prospective trial examining transoral resection of HPV-positive oropharyngeal cancer, followed by adjuvant therapy depending on pathological features. Low risk patients are managed by surveillance, high risk patients receive concurrent chemoradiotherapy, and intermediate risk patients (defined as < 1 mm extracapsular spread, 2 – 4 positive lymph nodes, or perineural/lymphovascular invasion) are randomized to 50 or 60 Gy postoperatively. The study requires 377 patients to achieve significance with 2-year progression free survival as endpoint [ ].
The ORATOR trial is a phase-II trial comparing TORS-based treatment to radiation with or without chemotherapy in patients with HPV-positive and HPV-negative oropharyngeal carcinoma, but with a 1-year quality of life as primary endpoint [ ].
Transoral robotic surgery (TORS) for squamous cell carcinoma of the posterior pharyngeal wall
Posterior pharyngeal wall squamous cell carcinoma (SCC) is rare and associated with poor prognosis. Overall 5-year survival is reported to lie between 3% and 32% [ ].
The relative rarity of these tumors and the heterogeneity in treatment modalities make the literature difficult to interpret, and treatment consensus is non-existent to date. Surgical therapy has long been neglected in favor of radiotherapy or chemoradiotherapy, considering the difficulty of access and reconstruction in this buried region. However, recent developments in transoral resection techniques and forearm free-flap reconstruction have given surgery a renewed role in the treatment of these cancers [ ].
Transoral robot-assisted surgery (TORS) using the da Vinci® system (Intuitive Surgical, Sunnyvale, CA) is a novel treatment option that has multiple advantages compared to conventional surgical techniques such as cervicotomy or transoral CO 2 laser surgery. Absence of communication between the resection area and the neck, better visualization of the resection area and the possibility of 4-handed surgery are all key advantages.
Since 2009, the French Head and Neck Robot-Assisted Surgery Group ( Groupe français de chirurgie robotique cervico-faciale ) has brought together the university hospital teams involved in TORS [ , ]. These teams use TORS for the treatment of selected patients presenting with posterior pharyngeal wall SCC. As there are to date no publications on the topic, the present article presents the oncologic and functional results obtained by the group using this innovative endoscopic approach.
Materials and methods
A retrospective multicenter study in 7 French university hospitals was approved by the institutional review board; all data were rendered anonymous. The patients included all had posterior pharyngeal wall SCC, of whatever T and N stages, and had undergone curative TORS between September 2009 and November 2013. All patients underwent primary tumor resection by TORS using the da Vinci® system, with patient positioning and resection as described by Weinstein [ , ]. The objective of the procedure was systematic complete en-bloc tumor resection, with surgical margins greater than 5 mm. Patients presenting with remote metastases or a secondary tumor site at the time of treatment were excluded.
Twenty-three patients were included: 7 women (30%) and 16 men (70%). Mean age at diagnosis was 62 years. Twelve patients had history of head and neck cancer (other than in the posterior pharyngeal wall), and had received initial curative treatment, including cervical irradiation.
Ten patients had stage cT1 tumor, 9 had stage cT2, and 4 had stage cT3. Three patients had lymph node involvement: 1 stage cN1, 1 stage cN2a and 1 stage cN2b.
The following treatments were performed: 8 patients were exclusively treated with TORS; 8 had TORS and neck dissection; 5 had TORS and neck dissection followed by adjuvant radiotherapy; and 2 had TORS followed by adjuvant radiotherapy. The postoperative indications for radiotherapy were presence of invaded or limited (< 3 mm) resection margins (6 patients), and pT3 tumor stage (1 patient). Eleven patients underwent neck dissection in the same step as TORS. For 2 patients, neck dissection was performed after TORS. Resection consisted in complete mucosal and muscular resection up to the prevertebral fascia for 19 patients, and mucosectomy associated to partial pharyngeal constrictor muscle resection for 4 patients. Reconstruction procedures were decided on by each team, based on resection size, adjacent tissue trophicity and history of cervical radiotherapy. First-line curative treatment without reconstruction was implemented in 13 patients. Biological glue was used in 8 patients. Forearm free-flap reconstruction was used in 1 patient, and 1 patient had a local mucosal flap.
Mean specimen size was 3.2 cm on the longest axis (range, 1.2–8 cm). Eleven patients had a stage pT1 lesion, 10 stage pT2 and 2 stage pT3. Histological analysis showed clear > 3 mm margins for 6 patients and clear < 3 mm margins for 13. Margins were invaded in 4 patients (17%), who underwent intraoperative re-excision; ultimately, only 1 of these patients had microscopic invasion after re-excision.
Mean follow-up was 27 months (range, 5–55 months), during which 6 patients had disease progression. There were 2 recurrences, exclusively confined to the tumor bed, 1 tumor recurrence with cervical lymph node involvement, and 3 cases of remote metastasis. Eight patients died during follow-up: 2 deaths were directly related to recurrence of the posterior pharyngeal wall lesion; 2 were related to a second primary tumor in the esophagus; and 4 deaths were due to comorbidity (1 spondylodiscitis, 1 renal cancer, 1 myocardial infarction and 1 of unknown cause). Two-year overall survival was 59%, and 2-year recurrence-free survival 56%.
There were no intraoperative complications. During the mean hospital stay of 12.7 days (range, 4–35 days), no cervical fistulae or hemorrhage occurred. Two patients developed pneumonia, implicating pharyngeal salivary stasis, resolving with antibiotic therapy. In the late postoperative period, 2 patients with mucosal and muscular resection up to the prevertebral fascia without reconstruction developed spondylodiscitis presenting as persistent neck pain, treated by prolonged intravenous antibiotics. Outcome was favorable for the patient with no history of cervical radiotherapy; the second patient, with history of cervical irradiation, died from a vertebral fracture with spinal cord transection, despite long-term use of a rigid neck brace.
Management of the respiratory tract
Two preventative tracheostomies were performed intraoperatively, due to particularly wide resection and unfavorable anatomical conditions for the respiratory tract. One of the two patients had history of cervical radiotherapy. The tracheostomy was removed at 13 and 90 days respectively. A third patient already had a permanent tracheostomy, following previous epilaryngeal (supraglottic) squamous cell carcinoma, treated by partial laryngeal surgery and radiotherapy, complicated by laryngeal stenosis.
Management of the digestive tract
Nineteen patients (82%) received postoperative enteral feeding. A nasogastric tube was placed in 15 patients, for a mean 22.3 days. Four patients had gastrostomy, permanent in all cases. One of these patients already had a long-term gastrostomy prior to TORS, due to pharyngeal stenosis secondary to surgery and radiotherapy for vallecular carcinoma. A second patient was never weaned from the gastrostomy, due to early local recurrence of incurable posterior pharyngeal wall cancer. Two patients were never able to recover sufficient oral feeding, due to wide pharyngeal resection for stage pT3 posterior pharyngeal wall carcinomas, and developed significant postoperative dysphagia; both had had previous cervical irradiation.
In the series as a whole, postoperative nutritional difficulties were prominent and often significant, most significantly in patients with previous cervical radiotherapy, where the duration of enteral feeding was 28 versus 12 days (p=0.01).
This study is the first to present oncological results and functional outcomes for patients treated by TORS for posterior pharyngeal wall SCC. It was a retrospective descriptive study, with certain limitations. The cohort of patients was small and heterogeneous, with unequal follow-up periods. However, compared to literature data for these rare tumors, this multicenter study constitutes an important series of patients operated on endoscopically for posterior pharyngeal wall SCC.
The theoretic advantage of this robot-assisted approach is to enable rapid, less invasive surgery and a clear tumor margins, for tumors with difficult access for classical endoscopic surgery. In 20 of the 23 cases, this surgical approach was deemed actually easier and faster, thanks to the technical advantages of the da Vinci® robotic system: 3D endoscopic vision to 0° or 30°, 2- or 4-instrument endoscopic surgery, and precise movement and tremor-filtration of the tele-manipulated instruments. However, it was found that this technological gain did not entail improved oncological or functional prognosis for the disease. Moreover, overall 2-year survival was 59%, with a significant difference between the pT1 and pT2-T3 groups. Although most studies of posterior pharyngeal wall carcinoma associated different pharyngeal locations and different stages, which makes comparison difficult, the present results seem nevertheless comparable to those in the literature. Five-year survival is between 21% and 48% with conventional surgery [ , ], and 36% with transoral CO 2 laser surgery, all stages combined [ ]. Fifty percent 3-year survival with radiotherapy was reported by Copper et al. [ ].
The functional results showed that TORS in this location induced significant dysphagia. In the immediate postoperative period, all patients showed severe dysphagia, commonly with odynophagia and/or stasis of food and saliva, due to in food bolus propulsion defect. These disorders required systematic placement of nasogastric tubes or gastrostomies. Four of our patients were permanently on enteral feeding, directly implicating the TORS resection in 2 cases. Extended sacrifice of the inferior pharyngeal constrictor muscles has negative functional consequences that must be anticipated and consented to by the patient. Definitive enteral feeding may be required, especially in case of previous cervical radiotherapy. Postoperative swallowing rehabilitation is also essential.
TORS offers a solution for posterior pharyngeal wall SCC. It provides an alternative to medical treatment for early stage T1 lesions. First-line use is debatable due to the associated postoperative nutritional difficulties. Indications are as an option for curative treatment in previously irradiated patients, where it is fairly safe and feasible.