Richmon et al. assessed the factors that contributed to length of stay after TORS in 91 patients [9]. Twelve percent of patients in their cohort experienced a complication. Postoperative bleeding occurred in 7% of patients with two patients having recurrent postoperative bleeding. Nearly all (94%) of the complications occurred in the first postoperative week with 38% of the complications occurring within 24 h of surgery. Asher et al. examined factors that contributed to bleeding after TORS in 147 patients [16]. They reported 11 (7.5%) patients with postoperative bleeding at a mean occurrence of 11.1 ± 9.2 days after surgery. The majority (82%) of these bleeds required management in the operating room. Another large study of 293 TORS procedures reported an average time to onset of bleeding being 7.3 days postoperatively (range 0–18 days) [7]. Glazer et al. reported all major postoperative bleeding occurred within 10 days [12]. Pollei et al. concluded that the greatest bleeding risk is present from postoperative day 7 to 14 [17]. The mean postoperative day for bleeding was day 10 with 83.6% of those bleeds occurring within 2 weeks of surgery. Thus, there seems to be a bimodal distribution of bleeding similar to what is observed in patients after tonsillectomy (Fig. 18.1). Patients should be educated about the risk of bleeding with TORS as well as the most likely times for bleeding.

Although there is a wide range of complication rates between studies, the rate of postoperative bleeding appears greater in patients undergoing TORS for malignancy compared to patients undergoing TORS for benign indications (Tables 18.2 and 18.3).

In the USFDA indication trial for TORS with the da Vinci Surgical System, the postoperative bleeding rate was 7.3% (2.8% requiring return to operating room) among 177 patient treated at 3 institutions [11]. Vergez et al. reviewed 130 patients undergoing TORS with 93% having a diagnosis of malignancy. The postoperative bleeding rate was 11.5% with 93% of patients treated in the operating room [20]. In comparison, a cohort of 243 patients undergoing TORS for sleep-related breathing disorders experienced a postoperative bleeding rate of 5% with only 34% managed in the operating room [6]. A study of 293 TORS procedures performed for benign disease experienced a postoperative bleeding rate of 4.1% [7]. A cohort of 166 patients who underwent TORS for OSA had a postoperative bleeding rate of 7.2% with 58% going to the operating room for cauterization; all but one patient had bleeding that originated from the tonsillar fossa [12]. Aubry et al. reported the highest rate of postoperative bleeding at 18.5% in 178 patients [15]. Interestingly, this group had a very high proportion of laryngeal and hypopharyngeal tumors (71%) suggesting that the decreased working space and more limited exposure of the larynx and hypopharynx can contribute to poorer hemostatic control.

Anticoagulation and antiplatelet therapy can affect postoperative bleeding rates and are usually withheld and/or bridged during the perioperative period. A review of 147 patients undergoing TORS revealed that 72% of the patients who had postoperative bleeding were on an antithrombotic medication for other comorbidities [16]. The postoperative bleeding rate in patients taking antithrombotic medication was significantly higher at 17% versus 3% (p = 0.02). They also noted that postoperative bleeding risk was greatest on postoperative days 7–14. A French review found that anticoagulation and/or antiplatelet therapy was a significant risk factor for postoperative bleeding (p < 0.05) [15]. Richmon et al. reported similar trends stating 50% of the patients who had postoperative bleeding were found to be on anticoagulation therapy [9]. Hoff et al. reported a postoperative bleeding rate of 4.1% and contributed two late postoperative bleeding episodes after re-initiation of clopidogrel or warfarin [7]. Patients who had anticoagulation at the time of surgery had higher rates of postoperative bleeding compared to patients not anticoagulated (13.5% vs. 8.1%) although this did not reach statistical significance (p = 0.2785) [7]. The authors in this study were so convinced of the association between bleeding after TORS and perioperative anticoagulation that they recommended withholding anticoagulation for 4 weeks postoperatively. However, at this time it remains unclear the optimal duration of time to withhold or bridge anticoagulation in patients treated using TORS.

The majority of bleeding after TORS is venous and self-limiting. However, potentially catastrophic arterial bleeding can occur after TORS. The incidence of life-threatening bleeding after TORS is unknown. No deaths from bleeding after TORS were reported in the USFDA indication trial [11]. However, by 2013, there were seven deaths from bleeding after TORS self-reported in a voluntary survey of TORS surgeons in the USA [13]. This is likely a gross underestimation, underscored by the fact that the response rate for the study was low and that the respondents were heavily weighted to high volume TORS surgeons. A variety of surgical techniques have been developed to minimize the risk of catastrophic bleeding after TORS, and some authors have advocated for routine transcervical ligation of feeding vessels to minimize or eliminate the risk of arterial bleeding after TORS.

Pollei et al. reviewed factors affecting bleeding rates in patients undergoing transoral oropharyngectomy by different approaches which included TORS, transoral laser microsurgery (TLM) and handheld cautery in 906 patients [17]. Of the 5.4% of patients with postoperative bleeding, 67% required operative intervention. In that retrospective study, prophylactic transcervical ligation of the external carotid system was performed during the primary surgery in 15.6% of patients. They reported no overall difference in bleeding rate after ligation compared to those patients who were not ligated (p = 0.21). Severe postoperative bleeding, defined as bleeding resulting in hypoxia/airway compromise requiring tracheostomy, cardiopulmonary arrest, or hemodynamic instability requiring of a blood transfusion, occurred less frequently in patients who had concurrent transcervical vessel ligation at 11.1% versus 25.8%. The difference was clinically meaningful but not statistically significant (p = 0.66). Vessel ligation was performed more frequently in patients with higher T classification (p = 0.002) since these patients were most likely to develop bleeding after TORS. So the authors recommended that patients with higher T classification should be considered for prophylactic transcervical ligation to decrease the rate and severity of bleeding after TORS.

More recently, Mandal et al. reviewed factors for postoperative bleeding after TORS in 224 patients with 185 cases performed for malignancy and 39 performed for benign indications [18]. 9.82% of these patients had varying degrees of bleeding after TORS. Prophylactic transcervical arterial ligation (9.1%) did not decrease overall postoperative bleeding rates when compared to the non-prophylactically ligated group (9.9%) (p = 1.00). There was a decreasing trend in frequency of severe bleeding after TORS, but this did not reach statistical significance (p = 0.70). Prior radiation therapy or chemoradiation therapy increased postoperative bleeding rates but not significantly (p = 0.09). Many experienced TORS surgeons routinely ligate branches of the ipsilateral external carotid system despite the paucity of data to date to support the effectiveness of the procedure. This is likely because the consequences of arterial bleeding after TORS can be dire and, although rare, may be preventable with a simple maneuver. A better understanding of the incidence and pathogenesis of catastrophic bleeding after TORS is needed to more clearly define the optimal strategies for prevention.

There are multiple cranial nerves that can be encountered performing TORS, especially for malignancy. The severity of injury can include neuropraxia, axonotmesis, and neurotmesis. In cases that involve malignancy, important nerves may be intentionally sacrificed for adequate resection. The glossopharyngeal, hypoglossal, and lingual nerves are all at risk during TORS. Every effort should be made to preserve these nerves as they are collectively instrumental in the function of swallowing. It is also important to remember that neurologic injuries can be either direct or indirect. Direct nerve injury (e.g., cutting the nerve) is far less common than indirect injury (e.g., nerve compression).

The glossopharyngeal nerve serves as the main afferent innervation for the tonsillar fossa and oropharynx. It descends from the jugular foramen and courses with the stylopharyngeus through the superior and middle constrictor muscles. The nerve can be visualized anterior and medial to these muscles [21]. A branch of this nerve is frequently encountered during TORS for tonsillar malignancy as it courses between the stylopharyngeus and styloglossus muscles (Fig. 18.2). Sacrifice of this branch is often necessary to ensure an oncologic resection of cancers involving the inferior tonsil and or glossopharyngeal sulcus. The functional impact of sacrifice of a branch of the glossopharyngeal nerve during TORS has not been formally described but appears inconsequential in the context of soft tissue and mucosa loss.

In contrast, injury to the hypoglossal nerve during TORS can be functionally devastating. In well-selected TORS cases, the hypoglossal nerve is typically not at risk. However, an increased risk of injury is observed in patients with recurrent disease, a history of radiation treatment, and/or bulky primary tumors. Muscle movement of the ipsilateral tongue during electrocautery dissection can be an important, albeit traumatizing, signal of proximity to the nerve. It is also important to recognize that hypoglossal nerve injury can occur during placement of surgical clips to control or prevent bleeding. The lingual nerve is also at risk during TORS. Risk of direct injury to the lingual nerve is particularly relevant for cancers that extend anteriorly toward the floor of the mouth.

A critical understanding of the anatomy of the submandibular triangle from an “inside-out” perspective is paramount to avoiding injury to the hypoglossal and lingual nerves. Early recognition of the submandibular gland and posterior belly of the digastric muscle during TORS for cancers involving the glossopharyngeal sulcus can help avoid direct nerve injury. The hypoglossal nerve is at most risk during TORS as it passes over the hyoglossus and runs along the superior border of the hyoid bone, deep to the digastric and mylohyoid muscles [21]. The lingual nerve which gives afferent and taste sensation to the anterior two-thirds of the tongue can be found on the lateral surface of the styloglossus muscle [22].

Finally, the internal branch of the superior laryngeal nerve is at risk during TORS for supraglottic cancers. After piercing the thyrohyoid membrane, the internal branch of the superior laryngeal nerve provides afferent innervation for the supraglottic laryngeal mucosa [23]. It is involved with the cough reflex and aspiration prevention. This nerve travels in close proximity to the superior laryngeal artery which requires deliberate ligation during TORS of the larynx.

The incidence of significant neurologic injury after TORS is reportedly low. In a large survey study, temporary (<2 month) hypoglossal nerve injury occurred in 0.9% out of 2015 patients undergoing TORS, prolonged (>2 month) hypoglossal nerve injury occurred in 0.1%, and inadvertent lingual nerve injury occurred in 0.6% of cases [13]. Richmon et al. reported there were no hypoglossal nor lingual nerve palsies in 91 consecutive patients [9]. Weinstein et al. reported only 1 patient with tongue numbness out of 192 patients undergoing TORS for malignancy [11]. Many large retrospective studies have not reported nerve injuries. In contrast, Vicini et al. reported a hypogeusia rate of 14.2% in 243 TORS procedures for sleep-related disorders with all resolving within 8 months [6]. This likely represents indirect compression injury to the lingual nerve which may be underreported in other TORS series for malignancy. The risk of compression injury to the lingual nerve would seem proportional to the size of the tongue, duration and extent of retraction, as well as the surgical defect. Anecdotally, many patients undergoing TORS will have some extent of temporary sensation or taste change of their tongue. As with all risks, this should be communicated with patients preoperatively.

With swallowing being compromised from odynophagia and dysphagia, the risk of aspiration and subsequent pneumonia is increased after TORS. Easa et al. evaluated swallowing outcomes for 78 patients that underwent TORS for OSA [8]. Gastrografin fluoroscopy was performed in the first postoperative week with only 6% having signs of significant aspiration. These patients all were without any swallowing complaints within 3 months and had no resulting significant weight loss. There was also no significant correlation between the volumes of tissue removed and the incidence of aspiration. A large review of TORS for benign indications reported pneumonia occurring six times in a cohort of 285 patients [7].

The risk of aspiration and pneumonia is likely greatest after TORS for malignancy, although there is a wide range in the incidence reported. In the USFDA indication trial for TORS, there was a 2.8% rate of pneumonia with two out of these five patients developing life-threatening complications of acute respiratory distress syndrome and pneumothorax [11]. Chia et al. noted 1.1% rate of aspiration pneumonia out of survey study of 2015 patients [13]. In a systemic review, Hutcheson et al. reported an incidence of postoperative pneumonia ranging from 0% to 7% in patients following TORS for oropharyngeal malignancy [14]. Recently, Aubry et al. reported an aspiration pneumonia rate of 15.5% and found that higher T-stage (T3, T4) and laryngeal location of the primary tumor were significant risk factors (p < 0.05) [15]. These authors attributed the high rate of aspiration pneumonia to the high percentage of patients with laryngeal tumors (47.2%). The reporting of aspiration after TORS is likely linked to the extent and timing of investigation as some degree of laryngeal penetration is common on early swallowing studies after TORS. Aggressive management of pain with early and frequent speech and language pathology coaching are critical to preventing aspiration and pneumonia.