Figure. Orientation of transoral robotic surgical specimen for pathologic analysis. A, Surgical specimen inked and pinned to cork board. B, Transection of surgical specimen to assess adequacy of deep margin.
Weinstein GS, Quon H, Newman HJ, Chalian JA, Malloy K, Lin A, Desai A, Livolsi VA, Montone KT, Cohen KR, O’Malley BW. Transoral Robotic Surgery Alone for Oropharyngeal CancerAn Analysis of Local Control. Arch Otolaryngol Head Neck Surg. 2012;138(7):628-634. doi:10.1001/archoto.2012.1166
Author Affiliations: Departments of Otorhinolaryngology–Head and Neck Surgery (Drs Weinstein, Newman, Chalian, and O’Malley), Radiation Oncology (Dr Lin), Hematology-Oncology (Drs Desai and Cohen), and Pathology and Laboratory Medicine (Drs Livolsi and Montone), University of Pennsylvania Medical Center, Philadelphia; Department of Radiation Oncology and Molecular Radiation Science, Johns Hopkins University, Baltimore, Maryland (Dr Quon).
Objective To evaluate local control following transoral robotic surgery (TORS) with the da Vinci Surgical System (Intuitive Surgical Inc) as a single treatment modality for oropharyngeal squamous cell carcinoma (OSCC).
Design Prospective, single-center, observational study.
Setting Academic university health system and tertiary referral center.
Patients Thirty adults with previously untreated OSCC.
Intervention Transoral robotic surgery with staged neck dissection as indicated.
Main Outcome Measures Local control and margin status.
Results Thirty patients were enrolled with previously untreated OSCC and no prior head and neck radiation therapy. Follow-up duration was at least 18 months. At the time of diagnosis, 9 tumors were T1 (30%); 16 were T2 (53%); 4 were T3 (13%); and 1 was T4a (3%). The anatomic sites of these primary tumors were tonsil in 14 (47%), tongue base in 9 (30%), glossotonsillar sulcus in 3 (10%), soft palate in 3 (10%), and oropharyngeal wall in 1 (3%). There was only 1 patient (3%) who had a positive margin after primary resection; further resection achieved a final negative margin. Perineural invasion was noted in 3 tumors (10%). No patient received postoperative adjuvant therapy. At a mean follow-up of 2.7 years (range, 1.5-5.1 years), there was 1 patient with local failure (3%).
Conclusion As the only modality used for treatment of pathologically low-risk OSCCs, TORS provides high local control and is associated with low surgical morbidity.
Transoral robotic surgery (TORS), a novel group of techniques first developed at the University of Pennsylvania, has been defined as a surgical approach for benign and malignant lesions of the oral cavity and laryngopharynx. It involves placing a minimum of 3 arms of the da Vinci Surgical System (Intuitive Surgical Inc) into the oral cavity with oral retractors.1 The TORS en-bloc surgical approach is not only a primary modality of therapy, but it also provides important information regarding pathologic tumor attributes that is useful to determine the need for subsequent adjuvant therapy.2
To date, the main oncologic indication for TORS has been the minimally invasive transoral resection of oropharyngeal squamous cell carcinomas (OSCCs).2- 4 Local control has been exceptionally high in published series, as high as 98% in a recent report of patients with advanced OSCC treated with TORS.2 Nonetheless, a factor confounding interpretation of the true effectiveness of TORS on local control has been that most patients in these cohorts have also had multiple histologically confirmed positive lymph nodes or extracapsular nodal disease resulting in the need for postoperative radiation or chemoradiation therapy to treat the neck as well as the primary tumor bed.2 In 1 large series, for example, only 2% of patients had positive margins, but the presence of nodal metastases resulted in 85% of patients receiving postoperative irradiation to the neck and primary site with or without chemotherapy.2Quiz Ref IDThe use of irradiation as well as chemotherapy in a large number of cases calls into question whether the success at the primary site was related to the TORS procedure itself or the addition of postoperative radiation or chemoradiation therapy.
In an effort to discover whether TORS alone, without postoperative radiation or chemoradiation therapy, can provide effective local control for mucosal OSCC, we decided to study a cohort of patients from 2 consecutive TORS single-arm, prospective, observational trials performed at the University of Pennsylvania. Within both of these studies was a cohort of patients with previously untreated OSCC who underwent TORS alone. These patients were a subset of the world's first research and clinical program in Transoral Robotic Surgery (TORS) utilizing the da Vinci Surgical System.
The primary objective of the current study was to assess the local control rate for a series of patients with OSCC who were treated with TORS followed by staged neck dissection as indicated without postoperative radiation therapy or chemotherapy. Secondary end points included evaluation of the safety and efficacy of this approach.
Data were collected from a pool of 422 patients who were enrolled in 2 consecutive protocols investigating TORS. Both protocols were approved by the institutional review board of the Hospital of the University of Pennsylvania, Philadelphia. Both studies were prospective, single-arm, observational trials. The preoperative inclusion criteria for the 2 TORS studies were (1) patients aged at least 18 years at the time of treatment; (2) presence of indications for diagnostic or therapeutic surgical approaches for benign or malignant diseases of the oral cavity or laryngopharynx; and (3) written informed consent and/or consent waiver by the institutional review board.
The tumor-related indications for TORS resection included previously untreated, biopsy-proven OSCC of stage I, II, III, IVA, or IVB. The preoperative contraindications for the studies were (1) unexplained fever and/or untreated, active infection, (2) pregnancy, (3) previous head and neck surgery precluding transoral robotic procedures, and/or (4) the presence of medical conditions contraindicating general anesthesia or transoral surgical approaches. The intraoperative contraindication was inability to adequately visualize anatomy to the extent necessary to perform the diagnostic or therapeutic surgical approach transorally, as detailed previously.2
Quiz Ref IDOf the original pool of 422 patients who were enrolled in the 2 consecutive prospective TORS trials, 30 had previously untreated OSCCs and underwent TORS alone between August 2005 and February 2010 at the University of Pennsylvania. These patients are the focus of the current study. Patients were observed for a minimum of 18 months after the TORS procedure. The most recent clinic visit assessment was in September 2011. The surgical procedures for TORS have been previously described.4,5 The da Vinci Surgical System was used in all cases of TORS. Staged neck dissection was performed in 27 of 30 patients (90%). The surgical sites of all patients were allowed to heal by secondary intention or local tissue flaps, as indicated.
The importance of achieving negative margins warrants further description of the collaborative method used by the surgeon and surgical pathologist during and following TORS. At the University of Pennsylvania, carcinoma at the margin is considered a positive margin ; within 2 mm or less is considered a close margin ; and greater than 2 mm is considered a negative margin. An overview of how pathologic specimens were managed following TORS is described elsewhere.5
The approach to specimen orientation used for all TORS cases is for the surgeon to first evaluate the en-bloc resected specimen in the operating room both visually and by palpation. If the margins look or are suspected to be close or positive, the surgeon immediately returns to the patient and paints the tumor bed with topical methylene blue to provide orientation for the new margin once it is resected. The surgeon then performs a wide reresection and sutures the new wider resection over the initial specimen. If the margin then both feels and looks grossly negative, the surgeon hand carries each TORS specimen to the surgical pathology laboratory and, in the presence of the pathologist, inks all margins with permanent inks of various colors and draws a diagram of the resected specimen for the record. The specimen is pinned to cork and sectioned (Figure) (Video).
The mucosal and deep margins are then grossly assessed by the surgeon and the pathologist. If the mucosal and deep gross margins are believed to be widely negative, then no frozen section is performed, and the surgical case is concluded. If the gross margin is clearly close or positive, then no frozen section is performed, and the surgeon returns to the operating room, paints the tumor bed with topical methylene blue, and performs a wide re-resection of the entire area of the resection bed in question. Once again, the surgeon personally brings this final margin specimen to the pathology laboratory and shows the pathologist how the new margin is oriented relative to the main specimen. If there is then agreement that the margin is grossly negative, no frozen section is performed, and the case is concluded.
Quiz Ref IDFrozen sections are performed only when the margins are equivocal. This can happen, for instance, when fat, tonsillar tissue, or minor salivary glands have a texture and appearance similar to that of the cancer and either the surgeon or the pathologist believes that the margin needs frozen-section assessment. If the frozen section finding is negative, the case is concluded, and if it is positive the surgeon will return to the operating room and repeat the topical methylene blue staining and reresection technique.
The indications for adjuvant therapy with radiation or concurrent chemoradiation following TORS and the rationale for staged neck dissection have been previously described.2 At the University of Pennsylvania, we use established clinical and pathologic criteria, as outlined in the National Comprehensive Cancer Network Guidelines for Head and Neck Cancer6 or supported by the medical literature, for either adjuvant radiotherapy or concurrent chemoradiotherapy. Briefly, the indications for postoperative irradiation include multiple positive lymph nodes, clinically endophytic T3 tumors (as opposed to large exophytic T3 lesions with a superficial base), and T4 tumors. Our indications for postoperative chemoradiation include positive surgical margins and/or extracapsular spread of nodal disease.7 In addition, pathologic high-risk features that are considered relative indications for irradiation at the University of Pennsylvania include perineural invasion and lymphovascular invasion. In the case of relative indications for adjuvant treatment, the risks and benefits of treatment vs observation are discussed in detail with the patient. Quon et al8 have previously described the rationale and approach for postoperative irradiation following TORS.
Postoperative follow-up was standardized. The minimum follow-up during the first year after diagnosis was every 3 months; during the second year, every 4 months; during the third and fourth years, every 6 months; and during the fifth year, 1 annual visit. At every follow-up visit, radiologic images were obtained, including either positron emission tomography (PET) or computed tomography (CT) with contrast or magnetic resonance imaging of the neck with or without chest radiographs.
Although single-arm cohort trials may be subject to significant bias, numerous steps were taken to ensure that bias was minimized. All clinical investigators who participated in the TORS prospective trials at the University of Pennsylvania shared equipoise concerning the standard surgical and nonsurgical treatments vs TORS for the patient population under investigation. Therefore, in an effort to avoid surgical selection bias, we offered the option of participation to all patients in the Otorhinolaryngology–Head and Neck Surgery outpatient clinics who fit the eligibility criteria for the 2 University of Pennsylvania prospective TORS trials.
An additional source of bias in surgical trials is the potential for variability in surgical technique among contributing surgeons. All surgeons who contributed patients to this trial were trained in a standardized approach to each TORS procedure, and consensus was reached concerning the en-bloc surgical approach.
It is known that a potential confounder inherent to single-arm surgical cohort trials is that patients who choose to participate in a surgical study vs a nonsurgical study might be healthier overall. Patient preoperative functional status and overall health status were assessed using the Karnosfky Score and the Charlson Comorbidity Index, respectively, to ensure a measure of comparability of the results of this series with those of other nonsurgical series.9 Of note, 1 patient with a T1N0 SCC of the soft palate was lost to follow-up within 1 month of surgery and was excluded from the analysis.
From August 2005 to February 2010, a total of 30 consecutive patients at the University of Pennsylvania with stage I, II, III, or IV OSCC were treated with primary TORS alone without postoperative irradiation or chemotherapy. The minimum follow-up was 18 months, and the mean follow-up was 2.7 years. Tumor and patient characteristics of the 30 study patients are listed in Table 1. Clinical and pathologic TNM staging is presented in Table 2.
In 30 primary TORS procedures for OSCC, final pathologic evaluation revealed 10 cases (33%) that were pathologic node positive, and 1 case with a positive margin (3%), defined as tumor presence at the inked margin. This patient had focal positivity of carcinoma in situ, and therefore reexcision at the time of neck dissection was considered a better option than postoperative chemoradiation therapy. One patient was over the weight limit for any radiation machine in our region and therefore was observed postoperatively.
Under the treatment regimen of primary TORS and staged neck dissection without postoperative radiation, this cohort achieved local, regional, and distant disease control in 29 of 30 (97%), 27 of 30 (90%), and 30 of 30 (100%) cases, respectively, at a minimum follow-up of 18 months. Overall survival for this cohort at the time of last follow-up was 30 of 30 (100%), also at a minimum follow-up of 18 months.
The 1 local recurrence 3 months and 19 days after TORS occurred in a patient with a T2N0 tumor (staged rT4 at the time of recurrence because of invasion of the medial pterygoid muscle). This patient had negative margins following TORS, with pathologic analysis showing a poorly differentiated SCC with invasion to a depth of 0.3 cm and no perineural invasion, angiolymphatic invasion, or dysplasia at the margin. There was no regional or distant metastasis. At last follow-up 1 year and 7 months after chemoradiation for the recurrence, this patient's clinical examination findings and PET CT findings were negative, and full nutrition was being maintained without a feeding tube.
Three patients had a regional recurrence. One patient originally had a cT3N2b tongue base cancer that was staged at pT1N2b at histologic examination after the primary TORS. This patient underwent radical neck dissection after TORS and was found to have 16 positive nodes and extracapsular spread. The patient was advised to undergo postoperative chemoradiation, which he refused. His disease then recurred in both necks 1 year, 11 months after TORS and underwent salvage chemoradiation treatment. One patient with a cT2N0/pT2N0 glossotonsillar sulcus tumor who underwent I-IV selective neck dissection with no positive nodes found had regional recurrence in the ipsilateral high neck 4 months after TORS and was treated with chemoradiation. The third patient with neck recurrence had a cT3N1/pT2N2b tongue base carcinoma. and the nodes showed no extracapsular spread. The patient was advised to undergo adjuvant radiation therapy and refused. He then developed a contralateral neck recurrence and was treated with a neck dissection that revealed 2 positive nodes without extracapsular spread. He was again advised to undergo radiation therapy. He eventually received chemoradiation at another institution.
The average intraoperative blood loss was 88 mL (range, 10-500 mL). No patient needed a blood transfusion. The average surgical time for TORS was 84 minutes (range, 31-152 minutes). No cases were converted to open resections. The average hospital stay was 3.6 days (range, 1-7 days).
At last follow-up, all patients were taking full oral diet without a feeding tube. One patient underwent temporary tracheostomy owing to preoperative morbid obesity. One patient had mild nasopharyngeal insufficiency following TORS radical tonsillectomy.
Complications in the 30-day postoperative period included 1 case of temporomandibular joint capsulitis that resolved with anti-inflammatory medication; 1 case of postoperative seizure that resolved with medication; 1 case of postoperative bradycardia that resolved without intervention; 1 case of dyspepsia related to tube feedings requiring outpatient management; and 2 cases of minor bleeding episodes that resolved without sequelae. One patient had acute renal failure that resolved with medical treatment in 12 days, and 1 patient was hospitalized related to syncope and dehydration and discharged after 24 hours without further complications There were no intraoperative or perioperative deaths and no life-threatening bleeding incidents.
Quiz Ref IDThe primary objective of this study was to assess the local control rates in 30 patients with OSCC treated with TORS alone without adjuvant therapy. The key finding in this study was the high local control rates observed, with only 1 local recurrence. One patient in this series had positive margins (3%) that were rendered negative by additional primary site resection at the time of the staged neck dissection, obviating the need for postoperative chemoradiation therapy and thus avoiding the morbidity associated with chemoradiation therapy. Three patients had perineural invasion (10%), and 1 patient (3.3%) had lymphovascular invasion at the primary site (3%), each of which is considered a relative indication for postoperative irradiation at our institution. All 4 of these patients refused irradiation without local failure. Surprisingly, 16 of 30 patients had overall clinical stage 3 or 4 disease (53%) and had no local failures at the primary site despite the lack of adjuvant of therapy. In addition, our findings confirmed the findings of prior studies that the morbidity of TORS alone for oropharyngeal cancer is low because there was no requirement for permanent feeding tubes and no perioperative mortality.
While the results of this trial may be intriguing, the limitations warrant delineation. One limitation is in the experimental design: this was a single-arm, prospective, observational trial and not a randomized controlled trial (RCT). Von Elm et al10 have pointed out that RCTs cannot, by their nature, be used to answer all of the important medical questions and that, indeed, properly conducted observational studies play an important role. Cooper11 has also pointed out the inherent difficulties of performing RCTs to answer surgical questions, which likely contribute to the fact that very few surgical trials are randomized. In our opinion, the questions posed in the present study are best addressed in single-arm observational trials, with the caveat that rigorous criteria be used both in study design and reporting.10
One limitation of this study is the absence of human papillomavirus (HPV) and p16 analysis. The time period during which this patient population was accrued predates the consistent testing for HPV or its pathologic surrogate marker p16 at the University of Pennsylvania, and therefore these data are not reported. In a previous study, Cohen et al12 noted that in a series of 37 cases of oropharyngeal cancer treated with TORS, HPV positivity was found in 74%, with 34 patients being serotype 16 (92%). Interestingly, this study showed no difference in oncologic outcome based on HPV status. Given the similar patient pool in the current study, it is reasonable to postulate that a similar incidence of HPV or p16 positivity would be present in the current series. Marur et al13 have noted that patients with HPV-related oropharyngeal cancers tend to present at a younger age with fewer comorbidities and longer overall life expectancies than patients with smoking-related cancers. Quiz Ref IDCurrently at the University of Pennsylvania, patients with oropharyngeal carcinoma are counseled concerning the risks and benefits of surgical (including TORS, when indicated) and nonsurgical options without regard to HPV status. Given the finding of no feeding-tube dependency and excellent oncologic outcomes in the present cohort, it is reasonable to consider future research studies using a primary TORS approach as a means of therapeutic deintensification in the HPV-positive patient.
The use of postoperative radiotherapy (PORT) for cancers of the head and neck has evolved from the seminal observations by Fletcher and Evers14 and Hamberger et al15 that risk of neck disease relapse increases following open transcervical surgical approaches to the primary site, especially with tumor transection, and that use of elective and comprehensive neck irradiation reduces these relapses. These observations led some radiation oncologists to favor PORT even in the absence of adverse pathologic risk features following open transcervical surgical approaches to the primary site owing to concerns of potential tumor seeding during the surgery.16 It is now clear that in the setting of node-negative disease with no primary site adverse features, the risk of local-regional relapse with observation will be 10% or less after open transcervical primary resection.17
With other transoral surgical techniques (eg, transoral laser microsurgery [TLM]) that offer primary tumor resection without the risks of surgical seeding of the exposed neck compartments, the risk of local failure following TLM alone without irradiation for previously untreated oropharyngeal carcinomas, in patients with predominantly node-negative necks, has been reported to be 6.8%.18 The findings in our series suggests that TORS, even without adjuvant radiotherapy, may offers local control rates similar to if not better than those seen with TLM. If so, the higher control rate may be associated with greater confidence in the surgical margin assessment seen with en-bloc resection that lends itself to potentially more accurate pathologic evaluation when the surgeon provides clear orientation of the specimen. Moreover, the high local control rates observed also provide assurance that in more advanced nodal disease following primary site treatment with TORS, greater confidence can be attributed to a negative margin. This has implications for the design and dose of adjuvant irradiation to the primary site and suggests that the post-TORS primary site could be well treated with standard postoperative doses and fields that encompass the surgical bed rather than the original, and usually larger, tumor volume. Whether the primary site in the patient with significant nodal metastases could even be excluded from PORT in the absence of any primary site pathologic risk features is a subject of ongoing research.
In 2002, Parsons et al19 reviewed the available literature from the prior 30 years comparing outcomes of surgery vs radiation for oropharyngeal cancer and concluded that given the higher complication rate with surgery, most oropharyngeal cancers should be treated with radiation. While Parsons et al conducted an outstanding study, the surgical approaches analyzed in their series included only open transcervical or transmandibular primary resections, which of course was appropriate for the historical period they studied from 1970 to 2000, but those procedures are very different from transoral techniques in terms of morbidity. In fact, the inherent problems with open surgical approaches were the very catalyst that led to clinical research into minimally invasive approaches such as TLM.20
While TLM has been widely adopted for the management of early glottic cancer and found to be very useful for the transoral management of supraglottic SCC, by comparison there is a paucity of literature reporting outcomes for TLM in OSCC.21 Transoral robotic surgery is a natural extension of the important and seminal research in TLM.20,22 The primary malignant indications for TORS to date have been for tumors of the oropharynx and supraglottis.23 Therefore, in an era in which open approaches have been largely supplanted by minimally invasive surgical approaches such as TLM and TORS, arguments denouncing the role of surgery based on the findings of Parsons et al19 are no longer valid.24
A reasonable question is whether the results of a single-arm cohort trial from 1 institution are generalizable. Consistency in technique among the surgeons in this study was achieved by adherence to previously agreed-upon surgical and pathologic principles. Transoral robotic surgery is a novel group of specific surgical procedures that were designed to be both consistent and reproducible. In addition, an important part of developing TORS was the concurrent effort to determine which tumors and sites would be best suited for the new procedures. In our opinion, the best way to ensure that other surgeons can reproduce our excellent outcomes is through our concise and clear articulation of the techniques via standardized and validated training. Ideally, training for TORS as for any new surgical technique should include an emphasis on the indications and contraindications for applying the surgery, the actual details of the surgical technique used to achieve en-bloc resection, and the proper orientation techniques for precise pathologic analysis of the resected specimen.
The impetus for studying this subset of TORS patients was to answer the question “are successful oncologic outcomes at the primary site related to the actual TORS resection, or are they a function of combining TORS with postoperative adjuvant therapy?” While most patients in this study had advanced clinical stage, almost all of them were found to have favorable pathologic features as well as pathologic down-staging at both the primary site and the neck (Table 2), which allowed deintensification of the overall treatment approach compared with, for example, definitive chemoradiation. In conclusion, while these data support TORS alone without postoperative irradiation for oropharyngeal carcinomas with favorable pathologic features, further investigation would be valuable to allow these results to be generalized. We strongly encourage other investigators to perform high-quality, single-arm, prospective, observational trials and encourage funding of multi-institutional phase 2 trials to confirm generalizability of the TORS techniques and outcomes.
Correspondence: Gregory S. Weinstein, MD, Department of Otorhinolaryngology–Head and Neck Surgery, 3400 Spruce St, 5 Ravdin, University of Pennsylvania Health System, Philadelphia, PA 19104 (email@example.com).
Submitted for Publication: March 26, 2011; final revision received March 11, 2012; accepted April 19, 2012.
Author Contributions: Dr Montone had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Weinstein, Quon, Chalian, and O’Malley. Acquisition of data: Weinstein, Quon, Newman, Chalian, Livolsi, Montone, and O’Malley. Analysis and interpretation of data: Weinstein, Quon, Chalian, Malloy, Lin, Desai, Montone, Cohen, and O’Malley. Drafting of the manuscript: Weinstein, Newman, Montone, and O’Malley. Critical revision of the manuscript for important intellectual content: Weinstein, Quon, Chalian, Malloy, Lin, Desai, Livolsi, Montone, Cohen, and O’Malley. Statistical analysis: O’Malley. Obtained funding: O’Malley. Administrative, technical, and material support: Weinstein, Quon, Chalian, Malloy, Desai, Montone, and O’Malley. Study supervision: Malloy and O’Malley.
Financial Disclosure: Drs Weinstein and O’Malley served as training consultants for training for Intuitive Surgical Inc until November, 2011. Dr Quon received 1-time reimbursement for travel expenses from Intuitive Surgical Inc. Dr Newman received proctoring honoraria from Intuitive Surgical Inc.
Online-Only Material: The Video is available here .
Previous Presentation: This research was presented at the Annual Meeting of the American Head and Neck Society; April 27, 2011; Chicago, Illinois.