A Next-Generation Single-Port Robotic Surgical System for Transoral Robotic Surgery

Key Points

Question

What is the value of a next-generation single-port robotic surgical system for transoral resection of oropharyngeal cancer?

Findings

In an analysis of 2 nonrandomized clinical trials that included 47 adults in the United States and Hong Kong who underwent single-port transoral robotic surgery for tumors of the oropharynx, transoral resection was performed safely with no intraoperative complications or device-related adverse events. Blood loss was minimal and surgical margins were free of disease in 97% of patients.

Meaning

Use of a single-port surgical robot appears to be safe, feasible, and effective for patients with oropharyngeal carcinoma.

Abstract

Importance

Transoral endoscopic head and neck surgery now plays an important role in the multidisciplinary management of oropharyngeal carcinoma. Previous generations of robotic surgical systems used a multiport system with a rigid stereo-endoscope and 2 wristed instruments that facilitated transoral robotic surgery.

Objective

To evaluate a new single-port robotic surgical system in head and neck surgery prospectively through concurrent nonrandomized clinical trials.

Design, Setting, and Participants

Two prospective clinical trials were conducted from December 16, 2016, to December 26, 2017, to assess the safety, feasibility, and performance of a flexible single-port robotic surgical system in 4 institutions, including 3 in the United States and 1 in Hong Kong. A total of 47 patients with tumors of the oropharynx were enrolled and underwent surgery. All patients were classified as having American Society of Anesthesiologists class I to III status and Eastern Cooperative Oncology Group status 0 to 1. An initial cohort of 7 patients underwent staging and endoscopic procedures for benign disease. The remaining 40 patients all had malignant tumors of the oropharynx.

Main Outcomes and Measures

Safety was measured by the incidence of device-related serious adverse events. Feasibility and performance were measured by the conversion rate from the use of the single-port robotic surgical system to either open surgery or the use of any other transoral technology required to complete the planned procedure. Secondary end points of swallowing function and surgical margins were also measured.

Results

All 47 patients (8 women and 39 men; mean [SD] age, 61 [8] years) safely underwent transoral resection with the single-port robotic surgical system without conversion to open surgery, laser surgery, or multiport robotic surgery. There were no intraoperative complications or device-related serious adverse events. Mean (SD) estimated intraoperative blood loss per procedure was 15.4 (23.9) mL; no patients received a transfusion. Two patients underwent a planned tracheotomy owing to medical comorbidity (previous chemoradiotherapy; obesity and severe sleep apnea). Two patients (4%) had grade III or IV postoperative hemorrhage, requiring a return to the operating room; however, both patients had medical comorbidities requiring the use of antithrombotic medication. The incidence of positive margins for patients with oropharyngeal malignancy was 3% (1 of 40). Within 30 days, 45 patients (96%) were eating by mouth and without the need for a percutaneous endoscopic gastrostomy tube.

Conclusions and Relevance

This study describes the results of phase 2 clinical testing of a next-generation, robotic surgical system using a single-port architecture. The use of the device appears to be feasible, safe, and effective for transoral robotic surgery of oropharyngeal tumors.

Trial Registration

ClinicalTrials.gov identifiers: NCT03010813 and NCT03049280

This study evaluates a new single-port robotic surgical system for head and neck surgery prospectively through 2 concurrent nonrandomized clinical trials.

Introduction

Transoral endoscopic head and neck surgery now plays an important role in the multidisciplinary management of head and neck cancer.¹ Using either robotics² or carbon dioxide laser and a microscope,³ transoral surgery provides a minimally invasive approach to treat oropharyngeal carcinoma. First-generation robotic surgical systems used a rigid stereoendoscope and 2 instruments that facilitated transoral robotic surgery (TORS). A new single-port (SP) robotic system has been approved for use in genitourinary surgery⁴ using a unique architecture ideally suited for minimally invasive surgery and, in particular, natural orifice endoscopic surgery.

Single-port robotic surgery has shown promise in preclinical testing of transoral resection of the lateral oropharynx,⁵ tongue base,⁶ and more caudal laryngopharynx.^7,8 The configuration and mechanical capabilities of the SP system have been more thoroughly described elsewhere^4,5,6,7,8; briefly, the SP system uses 3 fully articulating 6-mm instruments and a 1.2-cm stereoendoscope camera, all deployed through a single port measuring 25 mm and individually articulated with 2 joggle joints, mimicking the human wrist and elbow within the surgical field (Figure 1 and Video). In laboratory testing and human cases in genitourinary surgery, the SP system appeared to reduce clutter and improve ergonomics and dexterity. As such, we tested this system for its application in head and neck surgery in a series of 2 multicenter concurrent phase 2 clinical trials.

Figure 1. Single-Port Robotic Surgical System.

The da Vinci single-port system is a robotic surgical system in 3 parts. The surgeon console and vision cart are similar to those used with the earlier Xi robotic system. However, the patient cart has a unique architecture, with 3 fully articulating 6-mm instruments and a 1.2-cm stereoendoscope camera, all deployed through a single port measuring 25 mm. Each of the 3 instruments individually articulates with 2 joggle joints, mimicking the human wrist and elbow within the surgical field.

Video. Left Lateral Oropharyngectomy Performed Using a Single-Port Robotic Surgical System.

Download video file ^{(86.1MB, mp4)}

Methods

The working hypothesis was that an SP robotic surgical system would be feasible and safe to use for TORS and may offer certain advantages compared with multiport rigid robotics. Thus, in 4 centers, including 3 in North America and 1 in Hong Kong, 2 prospective clinical trials were performed from December 16, 2016, to December 26, 2017, to assess the safety, feasibility, and efficacy of a flexible SP robotic surgical system. The binocular stereoscopic camera displays full 1080-pixel high-definition visualization of human anatomy and works at an optimal focal length of 6 to 8 cm for autofocus. Each individual surgical instrument arm can extend as far as 25 cm from the cannula and can be manipulated off the central axis a maximum distance of 4 cm. The relationship of each instrument arm to the other instruments are shown in real time on the surgeon console, using a graphical user interface called the navigator. This interface allows the surgeon to avoid more proximal collisions between each instrument. Each arm’s deployment off-axis can be offset to optimize visualization, to enhance traction-countertraction, and to minimize distal collisions of instrument tips. For cutting and coagulation, both a monopolar spatulated cautery tip and scissors can be used. For soft-tissue manipulation, a Maryland dissector and a fenestrated grasper are available; through both of these instruments, bipolar cautery can be used for hemostasis. An impedance-matched electrosurgical platform is used (Erbe-Elektromedizin-GmbH), rather than traditional wattage-based electrocautery.

These studies represented stage 2 surgical evaluations based on the Innovation, Development, Exploration, Assessment, Long-term Study (IDEAL) framework⁹ and were conducted in accordance with the Declaration of Helsinki,¹⁰ administered with approval of the institutional review boards from the Chinese University of Hong Kong, Stanford University, The University of Pennsylvania, and Sunbelt Adventist Health System/Florida Hospital. Patients provided written informed consent. The studies were registered on https://www.clinicaltrials.gov/ (Hong Kong: NCT03010813¹¹; United States: NCT03049280¹²) (Figure 2). Data were then pooled across institutions to evaluate safety, feasibility, and efficacy.

Figure 2. CONSORT Diagram.

eHNS indicates endoscopic head and neck surgery.

The primary objective of these concurrent studies was to evaluate whether TORS could be performed using an SP robotic surgical system. Two objective measures were evaluated: one for safety and another for feasibility. Safety was measured by the incidence of device-related serious adverse events. Feasibility and performance was measured as the conversion rate from the SP robotic system to either an open surgical approach or the use of any other transoral technique (laser or multiport robotic surgical system) required to complete the planned procedure. Secondary end points of swallowing function and surgical margins were also measured. For the US multicenter phase 2 confirmatory studies, number of evaluable patients was not based on a formal sample size calculation. Rather, the clinical trials were intended to provide a measure by which the performance and safety of the SP system in a clinical setting can be confirmed based on previous generation TORS and preclinical animal and cadaver testing.^4,5,6,7,8

Eligibility criteria were standardized across institutions. Patients older than 18 years, with previously untreated benign or malignant tumors of the oropharynx classified as T1-2 were eligible to participate. The tumor must be accessible and amenable to transoral resection. Previous diagnostic biopsy of the oropharynx or excisional lymph node biopsy or neck dissection were allowed. Patients were excluded from the study if the tumor approximated the main trunk or branches of the external carotid arterial system. Patients who required an antithrombotic medication (including but not limited to warfarin sodium, enoxaparin sodium, or clopidogrel bisulfate) that could not be stopped prior to surgery were also excluded. Prospective functional outcomes were measured using the MD Anderson Dysphagia Index (MDADI) outcomes instrument.¹³

All surgeons in the current study received training on the SP robotic surgical system through didactic exercises using porcine and human cadaveric dissection models, based on previously defined procedural steps.^5,6 Transoral endoscopic resection was performed en bloc and under general anesthesia as described elsewhere in detail.^5,6

Data collection included preoperative patient characteristics, operative time and ability to complete the robotic procedure, tumor location, histologic characteristics, stage and tissue margin status, perioperative complications, total length of hospital stay, and the rate of percutaneous endoscopic gastrostomy tube dependency. Swallowing assessment was performed prospectively as part of the study, using the MDADI at 3 time points: prior to surgery, 2 weeks later, and again at final follow-up, which ranged from 30 days in the Hong Kong trial to 7 weeks in the US studies.

In the United States, the multicenter study of pooled prospectively collected data was conducted at Stanford University, Florida Hospital Celebration (currently AdventHealth Celebration), and the University of Pennsylvania and was sponsored by Intuitive Surgical Inc as part of an investigative device exemption trial submitted to the US Food and Drug Administration for assessment of the da Vinci Single-Port Surgical System (Intuitive Surgical Inc) for TORS. In Hong Kong, an investigator-sponsored study was conducted as part of a multidisciplinary surgical evaluation of the robotic system. The authors had full control of study design, analysis of the data, and writing this article.

In the United States, all adverse events were reviewed by an independent data safety monitor (E.J.M.) to determine severity and potential association with the SP robotic surgical system, the TORS procedure, and neck dissection or association with each patient’s medical condition. Predefined data elements were included prior to study launch and were prospectively collected using an electronic data capture system. Study monitors were used to confirm abstracted clinical data against medical records at each institution. Data were analyzed using SAS software (SAS Institute Inc). P < .05 was considered significant. In this report, all continuous variables are expressed as mean (SD) and range, whereas discrete variables are presented as rates and proportions.

Results

A total of 47 patients with tumors of the oropharynx were enrolled and underwent surgery. Table 1 documents the age, sex, Eastern Cooperative Oncology Group status, and other demographic characteristics of the study populations. The study included 8 women and 39 men, with a mean (SD) age of 61 (8) years. All patients were classified as having American Society of Anesthesiologists class I to III status and Eastern Cooperative Oncology Group status 0 to 1. All patients safely completed treatment with TORS using the SP robotic surgical system. An initial cohort of 7 patients with benign or indeterminate disease underwent staging and/or endoscopic procedures. There were no serious adverse events, so the remaining 40 patients were evaluated using the SP system for surgical resection of T1-T2 malignant tumors of the oropharynx, including 38 patients with squamous cell carcinoma and 2 with minor salivary gland carcinoma.

Table 1. Demographic and Tumor Characteristics of 47 Patients.

Characteristic	Patients, No. (%)
	All (N = 47)	Hong Kong (n = 14)	Florida Hospital Celebration (n = 11)	Stanford University (n = 11)	University of Pennsylvania (n = 11)
Age, mean (SD), y	61.0 (8.3)	59.2 (9.8)	66 (8.1)	59 (7.6)	60 (5.1)
Sex
Male	39 (83)	11 (79)	6 (55)	11 (100)	11 (100)
Female	8 (17)	3 (21)	5 (45)	0	0
Race/ethnicity
African American	2 (4)	0	0	1 (9)	1 (9)
Asian	13 (28)	13 (93)	0	0	0
Latin American	1 (2)	0	0	1 (9)	0
White	31 (66)	1 (7)	11 (100)	9 (82)	10 (91)
ECOG status
0 (Fully active, able to carry on all predisease performance without restriction)	40 (85)	13 (93)	10 (91)	6 (55)	11 (100)
1 (Restricted in physically strenuous activity but ambulatory and able to carry out light work)	7 (15)	1 (7)	1 (9)	5 (45)	0
ASA status
I	7 (15)	5 (36)	0	2 (18)	0
II	28 (60)	7 (50)	6 (55)	7 (64)	8 (73)
III	12 (26)	2 (14)	5 (45)	2 (18)	3 (27)
Histologic characteristics
Benign	7 (15)	7 (50)	0	0	0
Malignant	40 (85)	7 (50)	11 (100)	11 (100)	11 (100)
Oropharyngeal subsite
Tonsil	23 (49)	5 (36)	7 (64)	6 (55)	5 (45)
Tongue base	16 (34)	9 (64)	2 (18)	1 (9)	4 (36)
Glossopharyngeal sulcus	6 (13)	0	1 (9)	4 (36)	1 (9)
Soft palate	2 (4)	0	1 (9)	0	1 (9)

Abbreviations: ASA, American Society of Anesthesiologists; ECOG, Eastern Cooperative Oncology Group.

Table 2 presents the final pathologic tumor staging.¹⁴ Two patients’ tumors were staged clinically and radiographically as cT2 tumors; after measurement, the tumors were staged pT3 on final histologic assessment. A total of 37 of the 40 patients with cancer (93%), underwent neck dissection, which was performed concurrently on the day of TORS in 18 patients (45%) under general anesthetic for the transoral procedure. The neck dissection was staged in 19 of 37 patients (51%). For patients undergoing a neck dissection, 33 (89%) had prophylactic ligation of the external carotid artery or its appropriate arterial branches. Three patients undergoing TORS without neck dissection had either a soft palate tumor (US trials) or previous radiotherapy to the neck (Hong Kong trial).

Table 2. Final Pathologic Tumor Staging for 40 Patients With Oropharyngeal Carcinoma.

TNM Staginga	N0	N1	N2	N2a	N2b	N3	Nx	Total
T1	1	5	1	2	8	1	2	20
T2	1	4	2	3	4	0	1	15
T3	0	0	0	1	1	0	0	2
Tx	0	0	1	1	1	0	0	3
Total	2	9	4	7	14	1	3	40

^aStaged according to the American Joint Committee on Cancer’s Cancer Staging Manual, 7th edition.¹⁴

Surgical outcomes after TORS with the SP system are presented in Table 3. There were no intraoperative complications related to the device or TORS. The instruments were routinely deployed without limitation by bony anatomy or the natural width of the oral cavity and oropharynx. For all 47 patients, mean estimated blood loss was 15 mL, with no intraoperative or perioperative transfusions. There were no dental or lip injuries or facial or corneal abrasions. Negative surgical margins were achieved in 39 of the 40 patients with cancer (98%) enrolled in the study; in other words, only 1 patient had evidence of microscopic tumor on final histologic assessment.

Table 3. Surgical Outcomes for 47 Patients.

Outcome	Patients, No. (%)
	All (N = 47)	Hong Kong (n = 14)	Florida Hospital Celebration (n = 11)	Stanford University (n = 11)	University of Pennsylvania (n = 11)
Estimated blood loss, mean (SD), mL	15.4 (23.9)	22.9 (32.9)	5.5 (5.2)	4.7 (1.0)	26.4 (27.5)
Transfusion rate
No transfusion	47 (100)	14 (100)	11 (100)	11 (100)	11 (100)
Perioperative tracheostomy
No	45 (96)	13 (93)	11 (100)	11 (100)	10 (91)
Yes	2 (4)	1 (7)	0	0	1 (9)
Long-term tracheostomy
No	47 (100)	14 (100)	11 (100)	11 (100)	11 (100)
Presence of nasogastric feeding tube (at last visit)
No	45 (96)	13 (93)	10 (91)	11 (100)	11 (100)
Yes	2 (4)	1 (7)	1 (9)	0	0
Surgical margins
Negative	39 (98)a	7 (100)	10 (91)	11 (100)	11 (100)
Positive	1 (3)a	0	1 (9)	0	0

^aFor 40 patients.

Two patients underwent planned tracheostomy owing to medical comorbidities. Indications for these 2 patients were severe obstructive sleep apnea requiring continuous positive airway pressure as well as coronary artery disease with previous stenting and coronary artery bypass grafting surgery in 1 patient and airway edema and fibrosis due to previous radiotherapy and concurrent chemotherapy in another patient.

For centers in the United States, the mean (SD) length of hospital stay was 4.2 (1.9) days (2.4 [1.2] days at Florida Hospital Celebration, 5.4 [1.8] days at Stanford University, and 4.9 [0.8] days at University of Pennsylvania). The difference in length of stay at the 3 participating institutions may be attributed to standard practices at these institutions.

Overall, patients tolerated TORS performed with an SP robotic surgical system. Adverse events were scored by the Common Terminology Criteria for Adverse Events, version 4¹⁵ and the Dindo-Clavien postoperative complications schema.¹⁶ There were no deaths or grade 5 serious adverse events. Fifteen patients reported 20 adverse events; 4 of these patients experienced 5 serious adverse events. Five patients experienced grade 1 adverse events including dysgeusia, dysphonia due to postintubation vocal cord hematoma, drug allergy to postoperative antibiotic, and dysphagia with difficulty managing secretions. A single patient sustained a fall when combining a prescribed hydrocodone dose with a marijuana edible prescribed by another clinician outside the treating institution. However, no laceration or concussion was sustained and the patient recovered without treatment. Five patients experienced grade 2 adverse events, including 2 episodes of oropharyngeal bleeding, controlled with transoral cauterization of the tonsillar fossa in an outpatient clinic setting; cough and pneumonia, not requiring admission; nasal regurgitation, and trismus. All resolved with conservative medical management.

Four patients developed serious adverse events. A 69-year-old patient presented with grade III dehydration from Clostridium difficile enterocolitis, likely due to the use of postoperative antibiotics. The enterocolitis resolved with intravenous antibiotics. A 70-year-old patient presented with postoperative delirium and developed pneumonia. Two patients (4%) developed grade III or IV postoperative hemorrhage. Both episodes appeared 1 week or more after surgery and appeared related to unanticipated use of postoperative antithrombotic medications for the management of medical comorbidities. A 78-year-old patient with a right-sided oropharyngeal tumor developed a left middle cerebral artery occlusion that resulted in stroke on postoperative day 5. He had previously undiagnosed carotid artery atherosclerosis and was treated with tissue plasminogen activator immediately after the stroke. Two days later, oropharyngeal hemorrhage developed, requiring control in the operating room. A 57-year-old man with obesity, obstructive sleep apnea, and coronary artery disease treated with previous stent placement and subsequent coronary artery bypass graft surgery developed bleeding, requiring readmission after resumption of oral clopidogrel on postoperative day 16. Both patients had undergone selective, prophylactic ligation of at-risk branches from the ipsilateral external carotid artery system.

Within 30 days, 45 patients (96%) had resumed an entirely oral diet. One patient was not able to complete postoperative MDADI evaluations. Another patient from the Hong Kong study who already depended on a percutaneous gastrotomy tube underwent TORS for a second primary oropharyngeal tumor after previous treatment with radiotherapy with concurrent chemotherapy for a nasopharynx cancer and remained dependent on a percutaneous gastrostomy tube.

In addition, for patients in the US study, a detailed prospective assessment of swallowing function was performed using the MDADI. As expected, the mean MDADI scores at the immediate postoperative visit on each subscale (global, emotional, physical, and functional) declined compared with preoperative baseline scores (Figure 3). However, the mean MDADI subscale scores increased at the last follow-up visit, suggesting a return to baseline preoperative function.

Figure 3. Swallowing Function as Measured by the MD Anderson Dysphagia Index (MDADI) Score.

Preoperative and postoperative swallowing function as measured by the MDADI. Curves for the global scale and the individual functional, emotional, and physical scales are presented.

Discussion

We present the results of prospective clinical trials to evaluate the role of a next-generation SP robotic surgical system for transoral endoscopic head and neck surgery of the oropharynx, across 4 institutions in the United States and Hong Kong. Given that transoral endoscopic head and neck surgery is playing a greater role in the multidisciplinary management of oropharyngeal cancer, surgical innovation might increase the number of patients eligible for transoral surgery and, thus, help reduce the morbidity and late toxic effects of treatment. As popularized by Weinstein and O’Malley,² TORS has provided a standardized surgical approach for the treatment of this disease, which was approved by the US Food and Drug Administration for use in 2009.¹⁷ However, some aspects of earlier systems warranted further development to improve usability in TORS. The earlier-generation multiport surgical robotic systems had been originally designed for cardiac and abdomino-pelvic surgery, not the intricate anatomy within the narrow confines of the head and neck. Thus, the early robotic surgical instrumentation, while safe and effective, was not ideally suited to the pharynx and larynx. Moreover, access to transoral surgical anatomy and exposure is sometimes limited. With first-generation robotics, TORS could be performed with only 2 instruments and a binocular stereoendoscope. This arrangement limited the surgeon’s ability to provide optimal traction-countertraction and to manipulate soft tissues as effortlessly as in an open surgical field. Finally, while the tips of the instruments can be placed in close proximity to each other, improving on other forms of transoral surgery, the shaft of the instruments progressively spread apart from each other as they exit the oral cavity, resulting in potential challenges maneuvering around the jaws and oral retractor. The latter results in an additional level of skill that the surgeon must master during the learning curve. Although further study is needed, it is possible that the smaller profile of the SP trocar at the level of the jaws and the retractor will remove a level of complexity to surgical placement of the robotic arms, both shortening the learning curve and resulting in fewer bedside adjustments of the robotic arms during surgery.

In these prospective clinical studies, this SP next-generation robotic surgical system also appears safe for use in TORS for oropharyngeal cancer. There were no serious adverse events related to the use of the SP robotic system. All patients underwent planned procedures without intraoperative complication and without conversion to first-generation TORS, transoral laser microsurgery, or open surgical resection. During the TORS resection, there was minimal blood loss and no patients required transfusions. Detailed swallowing assessments using the MDADI after TORS with the SP system showed steady progress within the first month after surgery (Figure 3).

Using this SP robotic system, surgeons were able to perform surgery using up to three 6-mm instrument arms and a flexible stereoendoscopic camera. In these prospective studies, there were no new or unexpected adverse events relating to TORS. Among patients without medical conditions requiring antithrombotic medications, there were no episodes of postoperative hemorrhage. However, 2 patients (4%) who required antithrombotic medications postoperatively for medical comorbidities developed grade IV hemorrhage requiring reoperation. This rate of hemorrhage is consistent with previous rates of postoperative hemorrhage after transoral endoscopic head and neck surgery—ranging from 6.5%¹⁸ to 8%¹⁹ in the largest TORS series. Our rate (4%) was lower than these larger series, but most patients underwent planned prophylactic ligation of the at-risk branches of the external carotid artery, suggesting an important role for this technique for all patients undergoing TORS.

For tumors of the oropharynx, the SP robotic surgical system appeared well suited to function-preserving transoral endoscopic head and neck surgery. With the possibility of using up to 3 instrument arms and a flexible camera through a single port, and a smaller form-factor, this system may overcome some of the spatial limitations of the current robotic systems. In addition, the SP system may facilitate new procedures, including transoral surgery of larynx, hypopharynx, and skull base. Although not evaluated here, this system might have the capability to perform neck dissection and thyroid surgery via retroauricular approaches and even transoral approaches to the thyroid. However, further study would be required, first in the anatomic laboratory and then in prospective clinical evaluation.

Limitations

Current limitations of the system include instrumentation that has not yet been customized for transoral endoscopic head and neck anatomy. For example, more precise soft-tissue cutting and hemostasis would be useful, such as a finer monopolar electrocautery needle-tip end-effector. Having an additional instrument arm might also increase the chance of collisions not only with the teeth or retractor, but also potentially between each instrument. However, in this new system, a novel navigator panel demonstrates relationships between these arms, allowing surgeons to deploy and offset the arms in 3-dimensional space, thus minimizing instrument collision. In addition, the ability to deliver carbon dioxide or other laser energy might also help with mucosal cutting and hemostasis of deep tissues. Finer grasping forceps with bipolar cautery would also be of significant value. For fine work in the larynx, instruments delivering both laryngeal basket grasping capability and fine microscissors will be essential. Finally, with such a promising platform for surgery, improvement in visualization seems warranted. Instruments to deliver a flexible camera arm with an angled lens (ie, an additional 30°), an ultrasonographic probe, fluorescence imaging,²⁰ or multispectral or hyperspectral²¹ imaging of surgical anatomy would facilitate more precise procedures in the complex anatomical spaces of the head and neck.

Conclusions

We report the results of phase 2 prospective clinical trials evaluating a next-generation robotic surgical system using an SP architecture for TORS. The use of the device appears feasible, safe, and effective for transoral endoscopic head and neck surgery of oropharyngeal tumors.