Predictors of swallow function after transoral surgery for locally advanced oropharyngeal cancer

Abstract

Objective:

Transoral surgery (TOS) for oropharyngeal carcinoma (OPC) is steadily becoming more routine. Expected post-treatment swallow function is a critical consideration for preoperative counseling. The objective of this study was to identify predictors of swallow dysfunction following TOS for advanced T-stage (T3-T4) OPC.

Methods:

A retrospective review from 1997 to 2016 at a single institution was performed. Eighty-two patients who underwent primary transoral resection of locally advanced OPCs with at least 1 year of postoperative follow-up were included.

The primary outcome measure was swallow function, as measured by the Functional Outcomes Swallowing Scale (FOSS) at 1 year postoperatively. Operative reports were reviewed, and the extent of resection and type of reconstruction were documented. Conjunctive consolidation was then performed to incorporate multiple variables and their impact on swallow function into a clinically meaningful classification system.

Results:

Fifty-six patients (68%) had acceptable swallowing at 1 year. T4 tumor stage and receipt of adjuvant chemoradiation therapy (CRT) were strongly associated with poor swallowing but did not reach statistical significance. Only base of tongue (BOT) resection ≥50% (OR 3.19, 95%CI 1.21–8.43) and older age (OR 1.06, 95%CI 1.00–1.12) were significantly associated. Utilizing T-stage, adjuvant CRT, and BOT resection, a conjunctive consolidation was performed to develop a classification system for swallow dysfunction at 1 year.

Conclusion:

This study provides risk stratification for swallow function at 1 year following primary transoral resection of locally advanced OPCs. BOT resection ≥50%, especially when coupled with T4 tumor stage or adjuvant CRT, was associated with poor long-term swallow outcomes.

Introduction

Transoral surgery (TOS) for resection of oropharyngeal carcinoma (OPC) presents a minimally invasive surgical approach to an often difficult to reach area. Compared with standard transcervical approaches to the oropharynx, TOS is an attractive option for select patients, with the potential to decrease surgical morbidity while maintaining oncologic outcomes.¹ TOS encompasses transoral laser microsurgery (TLM), transoral robotic surgery (TORS) which is steadily becoming more widely accepted and routine, and loupe magnification with monopolar cautery.

The current standard of care therapeutic options for advanced OPC are definitive chemoradiation therapy (CRT) or primary surgery with adjuvant radiation or chemoradiation. Pre-treatment counseling is a critical step in helping patients decide which treatment modality to pursue. Aside from the requisite discussion on oncologic outcomes, post-treatment swallow function is a significant factor to consider. Definitive CRT has already been shown to cause up to 40% of patients to experience significant dysphagia within 3 years.² In contrast, multiple studies have shown much lower rates of swallow dysfunction in patients after TOS, with chronic gastrostomy tube rates <10%.^3–5 Similar studies also note that extensive base of tongue (BOT) and soft palate (SP) resection or cranial nerve (hypoglossal, vagus) transection can lead to worse swallow function.^6–9 Increased extent of cervical lymphadenectomy has also been correlated with swallow dysfunction.¹⁰ However, most surgical reports do not include transorally resected advanced tumors (T)-stage T3-T4, nor delineate the association between the extent of the defect and the impact of reconstruction on swallow function. The objective of this study is to identify predictors of swallow function at 1 year following transoral resection of locally advanced OPC.

Materials and Methods

Setting and Subjects

All data collection and analyses were approved by the Institutional Review Board of Washington University School of Medicine. A retrospective review was performed of patients undergoing primary transoral resection of locally advanced (pathologic T3-T4) OPC at a single academic institution between 1997 and 2016. T3 and T4 tumor stages were determined by the American Joint Committee on Cancer (AJCC) 7^th edition guidelines.¹¹ Pathologic rather than clinical T-stage was utilized due to its closer correlation with post-surgical defect size. Primary transoral resection included TORS, TLM, or loupe magnification under direct visualization and encompassed a previously published cohort whose swallowing outcomes were investigated.^12,13 Patients ≥ 18 years-old who underwent definitive surgical transoral treatment (+/− hybrid lateral pharyngotomy if complete tumor extirpation was limited by transoral access)¹⁴ and were disease-free at 1 year postoperatively were included in the study. Criteria for exclusion were: prior diagnosis or treatment of head and neck cancer, primary nonsurgical treatment, histologic diagnosis other than squamous cell carcinoma, recurrence within 1 year of surgery, and tumor resection requiring conversion to an open transcervical mandibulotomy.

Data Collection

Collected data included demographics, comorbidities as measured by the Adult Comorbidity Evaluation-27 (ACE-27) score, tumor presentation and pathology, surgical details, adjuvant therapy, recurrences, and postoperative tracheostomy dependence.¹⁵ Each patient’s primary operative report was carefully reviewed by physicians (JG, MT, JR) familiar with transoral surgical anatomy who were blinded to swallow outcomes. The extent of resection of each oropharyngeal subsite (SP, lateral pharynx/tonsil, and BOT) was documented and graded by percent of resection. Extirpation of critical neurovascular, muscular, or cartilaginous structures was documented. Creation of a pharyngotomy or an exposed carotid artery, concurrently performed neck dissection, as well as detailed descriptions of the reconstructive maneuvers (local, regional or free flap) were also recorded. Pathology reports were reviewed for pathologic TNM staging, tumor histology, and p16 status.

Receipt of adjuvant therapy (radiation and/or chemotherapy) and the dosage and location of the radiation field (primary site and/or neck irradiation) were collected. Clinic notes were reviewed for timing of recurrent locoregional or distant disease or death. Pre- and postoperative gastrostomy tube use as well as presence of long-term postoperative tracheostomy (>1 year) were recorded.

Outcome Measures

The primary outcome measure was swallow function at 1 year postoperatively, as measured by the Functional Outcome Swallowing Scale (FOSS)¹⁶ (Table 1). The FOSS scale incorporates subjective and objective factors to create a nominal classification of swallow function from 0 to 5. Zero represents normal function without any symptoms, and 5 represents complete dependence on nonoral feeding. Patients with a FOSS score 0–2 had compensated swallow function with only episodic symptoms of dysphagia and were classified with acceptable function. Patients with a FOSS score 3–5 had decompensated swallow function with weight loss, recurrent aspiration, or nonoral feeding and were classified with poor function. Two physicians (JG, MT) reviewed speech pathology evaluations, gastrostomy tube status, and physician follow-up notes to determine FOSS scores by consensus. The division into “acceptable” and “poor” FOSS score subsets was adopted from a previously published study.¹³

Table 1.

Functional Outcome Swallowing Scale (FOSS)¹⁶

Stage	Symptoms
0	Normal function and asymptomatic.
1	Normal function with episodic or daily symptoms of dysphagia.
2	Compensated abnormal function manifested by considerable dietary modifications or prolonged mealtime (without weight loss or aspiration).
3	Decompensated abnormal function with weight loss of 10% of body weigh over 6 months owing to dysphagia; or daily cough, gagging, or aspiration during meals.
4	Severely decompensated abnormal function with weight loss of 10% of body weight over 6 months owing to dysphagia; or severe aspiration with bronchopulmonary complications. Nonoral feeding for most nutrition.
5	Nonoral feeding for all nutrition.

^Ɨ(Adapted version provided by Rich, et al.¹³)

Statistical Analysis

Statistical analysis was performed using SPSS Statistics 24 (IBM) and SAS 9.4 (SAS Institute). Basic descriptive statistics were used to define patient characteristics. Logistic regression analysis was utilized to detect any associated differences between the acceptable and poor swallow function groups.

Conjunctive consolidation was performed to incorporate multiple variables and stratify data into a clinically meaningful classification system.¹⁷ This statistical technique created a classification predictive of poor post-treatment swallow function. Odds ratios from a logistic regression model were used to assess strength of association between the novel classification system and outcomes after controlling for other potential risk factors. The concordance statistic (c-statistic) was calculated, which determines the probability that a random patient will experience the event in question and is used to assess discrimination of the novel staging system. A value < 0.5 reflects a poor model, and a value of 1 indicates a perfect fit.

Results

Descriptive Statistics

A total of 112 cases of advanced stage OPC undergoing TOS were evaluated, and 82 patients met eligibility criteria and were included in the study. Pathologic tumor stage was T3 in 45 patients (55%), and T4 in 37 patients (45%). Eight patients (10%) had no BOT resection, 45 patients (55%) had BOT resection >0% and <50%, and 29 patients (35%) had ≥50% BOT resection. Six patients (8%) received no adjuvant therapy, 31 (39%) received adjuvant radiation therapy only, and 42 (53%) received combined adjuvant CRT. Adjuvant therapy status for 3 patients was unknown. The radiation fields could not be determined in most patients.

At 1 year postoperatively, swallow function was acceptable (FOSS 0–2) in 56 patients (68%), and poor (FOSS 3–5) in 26 patients (32%). Thirteen patients, or one half of those with poor swallow function, were gastrostomy tube dependent. Patient demographics, clinical characteristics, and outcomes are demonstrated in Table 2.

Table 2.

Patient demographics, clinical characteristics, and swallow outcomes.

	All n(%)	Acceptable Swallowing n(%)	Poor Swallowing n(%)	OR Ɨ	95% CIǂ
Age (years)
Mean (SD)	59.4 (9.2)	58.0 (9.2)	62.5 (8.3)	1.06	1.00–1.12
Age (dichotomous)
Less than 60	45 (55)	34 (61)	11 (42)
60 or Greater	37 (45)	22 (39)	15 (58)	2.11	0.82–5.42
Gender
Male	74 (90)	50 (89)	24 (92)
Female	8 (10)	6 (11)	2 (8)	0.69	0.13–3.70
Smoking
Never	31 (38)	23 (41)	8 (30)
Past	33 (40)	24 (43)	9 (35)	1.08	0.36–3.27
Current	18 (22)	9 (16)	9 (35)	2.88	0.84–9.79
Smoking
Never	31 (38)	23 (41)	8 (31)
Past or Current Smoker	51 (62)	33 (59)	18 (69)	1.57	0.58–4.21
ACE
0 (None)	33 (40)	24 (43)	9 (35)
1 (Mild)	40 (49)	26 (46)	14 (54)	1.44	0.53–3.92
2 (Moderate)	7 (9)	4 (7)	3 (11)	2.00	0.37–10.75
3 (Severe)	2 (2)	2 (4)	0 (0)	NA	NA
ACE
None	33 (40)	24 (43)	9 (35)
Comorbidities Present	49 (60)	32 (57)	17 (65)	1.42	0.54–3.72
Site (Consolidated)
BOT	42 (51)	26 (46)	16 (62)
Tonsil	40 (49)	30 (54)	10 (38)	0.54	0.21–1.40
Tumor Histology
Well-differentiated	0 (0)	0 (0)	0 (0)
Moderately differentiated	9 (11)	4 (7)	5 (20)
Poorly differentiated	16 (20)	10 (19)	6 (24)	0.48	0.09–2.52
Non-keratinizing	48 (61)	36 (67)	12 (48)	0.27	0.06–1.16
Other	6 (8)	4 (7)	2 (8)	0.40	0.05–3.42
Missing	3
pT Pathologic Tumor Stage (AJCC 7th ed.)
T3	45 (55)	35 (63)	10 (38)
T4a, T4b	37 (45)	21 (37)	16 (62)	2.67	1.02–6.95
pN Pathologic Nodal Stage (AJCC 7th ed.)
N0	12 (15)	11 (20)	1 (4)
N1	7 (9)	5 (9)	2 (8)	4.40	0.32–60.61
N2a	9 (11)	8 (14)	1 (4)	1.38	0.07–25.43
N2b	30 (37)	21 (37)	9 (36)	4.71	0.53–42.17
N2c	21 (26)	9 (16)	12 (48)	14.67	1.59–135.32
N3	2 (2)	2 (4)	0 (0)	NA	NA
Missing	1
pStage Pathologic Stage (AJCC 7th ed.)
I, II, III	9 (11)	8 (15)	1 (4)
IVA, IVB	71 (89)	46 (85)	25 (96)	4.35	0.51–36.78
Missing	2
Type of Surgical Treatment
TLM	74 (90)	48 (86)	26 (100)
TORS	4 (5)	4 (7)	0 (0)	NA	NA
Loupes	4 (5)	4 (7)	0 (0)	NA	NA
BOT Resection
0–49% Resection	53 (65)	41 (73)	12 (46)
≥50% Resection	29 (35)	15 (27)	14 (54)	3.19	1.21–8.43
Lateral Pharynx/Tonsil Resection
None to Parapharyngeal Fat Exposure	73 (89)	49 (88)	24 (92)
Exposed Carotid	9 (11)	7 (12)	2 (8)	0.58	0.11–3.02
Soft Palate Resection
0–49% Resection	77 (94)	53 (95)	24 (92)
≥50% Resection	5 (6)	3 (5)	2 (8)	NA	NA
Neck Dissection
None	4 (5)	3 (6)	1 (4)
Ipsilateral	40 (49)	31 (55)	9 (35)	0.87	0.08–9.43
Bilateral	38 (46)	22 (39)	16 (61)	2.18	0.21–22.95
Neck Dissection
None or Ipsilateral	44 (54)	34 (61)	10 (38)
Bilateral	38 (46)	22 (39)	16 (62)	2.47	0.95–6.43
CNX Resection
None	81 (99)	56 (100)	25 (96)
Performed	1 (1)	0 (0)	1 (4)	NA	NA
CNXII Resection
None	73 (89)	51 (91)	22 (85)
Performed	9 (11)	5 (9)	4 (15)	NA	NA
Lingual Nerve Resection
None	71 (87)	49 (88)	22 (85)
Performed	11 (13)	7 (12)	4 (15)	1.27	0.34–4.80
Pterygoid Resection
None	68 (83)	44 (79)	24 (92)
Performed	14 (17)	12 (21)	2 (8)	0.31	0.06–1.48
Type of Reconstruction
None, Local Flap, or Other	75 (92)	52 (93)	23 (89)
Regional or Free Flap	7 (8)	4 (7)	3 (11)	NA	NA
Pharyngotomy
None	47 (57)	34 (61)	13 (50)
Performed	35 (43)	22 (39)	13 (50)	1.55	0.61–3.95
p16
Negative	5 (6)	4 (7)	1 (4)
Positive	74 (94)	51 (93)	23 (96)	1.80	0.19–17.05
Missing	3
Adjuvant Therapy
None or Radiation Only	37 (47)	29 (54)	8 (32)
Chemoradiation Therapy	42 (53)	25 (46)	17 (68)	2.47	0.91–6.68
Missing	3
Recurrence (Any)
None	60 (73)	43 (77)	17 (65)
Present	22 (27)	13 (23)	9 (35)	1.75	0.63–4.85
Recurrence (Local or Regional)
None or Distant	71 (87)	48 (86)	23 (89)
Local or Regional	11 (13)	8 (14)	3 (11)	0.78	0.19–3.23
Long-term Tracheostomy, 1 year follow-up
No	77 (94)	55 (98)	22 (85)
Yes	5 (6)	1 (2)	4 (15)	10.00	1.06–94.54
Trismus, 1 year follow-up
No	65 (79)	46 (82)	19 (73)
Yes	17 (21)	10 (18)	7 (27)	1.70	0.56–5.11

^ƗOR = Odds Ratio

^ǂCI = Confidence Interval

Logistic Regression

Patients with a T4 tumor stage had significantly greater risk of poor swallow function when compared to T3 (OR 2.67, 95% CI 1.02–6.95). Fifty percent or greater BOT resection compared with <50% resection was also significantly associated with poor swallow function (OR 3.19, 95% CI 1.21–8.43). Older age had a small but statistically significant effect (OR 1.06, 95% CI 1.00–1.12). Patients who received adjuvant CRT tended to have greater risk of swallow dysfunction compared to those who received adjuvant radiation only or no adjuvant therapy, although this effect was not statistically significant (OR 2.47, 95% CI 0.91–6.68). Extent of resection of lateral pharynx/tonsil, medial pterygoid resection, extent of neck dissection, and pharyngotomy did not statistically impact swallow outcomes. The following variables had too few numbers and were not analyzed: type of reconstruction, extent of SP resection, and vagus and hypoglossal nerve transection.

Conjunctive Consolidation

Due to the small number of patients, only three variables could be utilized in conjunctive consolidation analysis without risk of overfitting the model. The variables with the highest clinical significance and strongest statistical association were chosen for analysis and included: BOT resection ≥50%, T-stage, and receipt of adjuvant CRT. Utilizing these three factors, a conjunctive consolidation was performed to show their composite effects on swallow dysfunction at 1 year and stratify the outcomes. First, extent of BOT resection and T-stage were conjoined to create a three-category staging system with Groups A, B, and C (Table 3). Patients with T3 tumors and BOT resection <50% had the lowest rate of swallow dysfunction (6 of 31, 19%) and formed Group A. Patients with T3 tumors and BOT resection ≥50% or T4 tumors and BOT resection <50% comprised Group B and had intermediate risk of swallow dysfunction (10 of 36, 28%). The highest rate of swallow dysfunction (10 of 15, 67%) formed Group C and was comprised of patients with T4 tumors and ≥50% BOT resection. The new classification incorporating BOT resection and T-stage information had good discriminative power, with a c-statistic of 0.68.

Table 3.

Rates of poor swallow function at 1 year postoperatively in conjunction with base of tongue (BOT) resection and pathologic tumor stage.

The third variable (adjuvant CRT) was then conjoined with the new Group ABC consolidation to create the final combined classification system for predicting swallow dysfunction at 1 year (Table 4). Patients in Group A or Group B with no adjuvant CRT, or those in Group A with adjuvant CRT defined Class I and had the lowest rate of swallow dysfunction (8 of 46, 17%). Patients in Group B with adjuvant CRT or Group C with no adjuvant CRT comprised Class II and had intermediate rates of swallow dysfunction (11 of 26, 42%). The highest rate of swallow dysfunction (6 of 7, 86%) formed Class III and was comprised of patients in Group C with adjuvant CRT. Thus, the final classification system demonstrated rates of swallow dysfunction as follows: Class I - 17%, Class II - 42%, and Class III - 86%. The c-statistic for the new classification system was 0.72.

Table 4.

Rates of poor swallow function at 1 year postoperatively in conjunction with BOT resection, pathologic tumor stage, and adjuvant chemoradiation therapy (CRT).

Discussion

Traditionally, surgical approaches to advanced oropharynx tumors require a mandibulotomy or large transcervical pharyngotomy. These techniques often require reconstruction and tracheotomy and have high functional morbidity due to disruption of musculoskeletal structures responsible for the oral and pharyngeal phases of swallowing. TOS has the potential to decrease surgical morbidity while maintaining oncologic outcomes.¹

This study demonstrated that patients undergoing primary transoral resection of locally advanced OPC had varying levels of swallow dysfunction, predicted by the amount of BOT resection, pathologic T-stage, and receipt of adjuvant CRT. Yet, most patients (56 patients, 68%) had acceptable swallow outcomes at 1 year postoperatively. Of the 26 patients (32%) with poor swallow outcomes, one half (13 patients, 16%) were gastrostomy tube dependent, which is within the wide range of what is reported in the literature for T3-T4 OPC (9 to 36%).^3,18,19

In bivariate analysis, patients with pathologic T-stage 4 were more likely to have poor swallow function than those with T-stage 3 (OR 2.67). This finding is clinically intuitive, and previous studies have shown similar outcomes with increasing T-stage.^13,19 For example, Dziegielewski et al. published a 27 times increased risk of requiring gastrostomy tube feedings in T4 versus T3 OPC removed via TORS.³

In this study, patients with ≥50% BOT resection were over 3 times more likely to have poor swallow function (OR 3.19). Extent of resection of SP or lateral pharynx/tonsil was not significantly impactful. The BOT is critical to the pharyngeal phase of swallowing, contributing to bolus propulsion and airway protection through epiglottic tilt.²⁰ As a result, it is reasonable to expect extensive BOT resection or CRT to lead to worse swallow outcomes, and multiple studies support this notion. Pauloski et al. examined 144 patients after primary surgery +/− adjuvant therapy for oral cavity cancer or OPC.⁶ Oropharyngeal swallow efficiency (OPSE) was negatively correlated with the percent of tongue base resected. In a retrospective review of advanced stage OPC treated with definitive CRT, Shiley et al. found a significantly higher rate of gastrostomy tube dependence in patients with BOT versus tonsil tumors (67% vs 25%).²¹

The patients in the current study all had advanced stage disease and were therefore recommended to undergo at least adjuvant radiation. It is well-established that any increase in radiation dosage may significantly worsen swallow function via post-radiation fibrosis of swallowing organs, specifically the pharyngeal constrictors and the glottic and supraglottic larynx.^22–26 The dose-dependent increase in potential side effects (i.e. mucositis, fibrosis, stricture) is an important point for pretreatment counseling, as the risk is considerably higher with definitive versus adjuvant radiation doses.

Forty-two patients (53%) in the study underwent additional concurrent adjuvant chemotherapy due to the presence of positive margins or extranodal extension. These patients had a 2.5 times greater risk of swallow dysfunction when compared to those who received radiation only or no adjuvant therapy, although it was not statistically significant (OR 2.47, 95% CI 0.91–6.68). The synergistic impact on swallow dysfunction from CRT is well-supported in the literature. A Surveillance Epidemiology and End Results (SEER) analysis by Francis et al. demonstrated a 2.7 odds ratio of dysphagia among patients receiving definitive CRT versus those undergoing surgery alone, compared with a 2.0 odds ratio among those undergoing surgery with adjuvant radiation versus surgery alone.² More et al. performed a prospective study comparing swallow function of patients with advanced stage oropharynx and supraglottis cancers treated with TORS and adjuvant therapy versus primary CRT.²⁷ Long term swallow-related quality of life, as measured by the MD Anderson Dysphagia Inventory (MDADI), was almost two times higher in patients with T3 tumors treated with TORS and adjuvant therapy versus primary CRT. Additionally, 60% of those patients undergoing TORS had adjuvant CRT, and they still reported better swallow scores. The above considerations underscore the critical importance of ongoing de-escalation studies in HPV-mediated OPC in which radiotherapy dose/treatment volume and the use of adjuvant CRT are under investigation.²⁸

Overall, only 17 patients (21%) underwent a local flap, and 7 patients (9%) underwent a regional or free flap reconstruction. Due to the low number of patients undergoing flap reconstruction, no conclusions could be made regarding the impact of reconstructive efforts on swallow function. Of note, 22 out of 29 (76%) of those patients with ≥50% BOT resection did not have any type of reconstruction. Though the data is sparse, other studies have shown variable swallow function with flap reconstruction for defects involving ≥50% BOT resection. Rieger et al. followed a group of 32 patients with OPC requiring >50% BOT resection followed by radial forearm free flap (RFFF) reconstruction.²⁹ At 1 year postoperatively, they showed good swallow outcomes, with only 14% (3/21) of patients requiring use of a gastrostomy tube. Seikaly et al. prospectively evaluated swallow function after RFFF reconstruction as well. Out of 9 patients with T3-T4 disease and ≥50% BOT resection, 3 (33%) had aspiration or required a gastrostomy tube.³⁰ In a cross-sectional study by Winter et al., 5/9 (56%) T3-T4 OPC patients with BOT free flap reconstruction reported poor swallow-related quality of life.³¹ In our study, 14/29 patients (48%) with ≥50% BOT resection had poor swallow function, which is on the higher end of published data. This data may be due to a lower number of flap reconstructions in our population.

Interestingly, there were multiple factors that were not statistically significant in bivariate analysis that intuitively would be thought to impact swallow function, namely extent of lateral pharynx/tonsil resection, resection of pterygoid musculature, tracheostomy dependence, extent of neck dissection, and limited pharyngotomy. The frequency of SP and cranial nerve (hypoglossal, vagus) resection was too low to make inferences on the impact on swallow function.

This paper presents a novel classification system that consolidates three predictive factors of poor swallow function that have also been supported by previously published data. Namely tumor stage, extent of BOT resection, and adjuvant CRT are combined to stratify patients into low, intermediate, or high risk of post-treatment swallow dysfunction. The low risk group included: patients with T3 tumors and <50% BOT resection, T3 tumors and ≥50% BOT resection without adjuvant CRT, or T4 tumors and <50% BOT resection without adjuvant CRT. The intermediate risk group included: patients with T3 tumors and ≥50% BOT resection with adjuvant CRT, T4 tumors and <50% BOT resection with adjuvant CRT, or T4 tumors and ≥50% BOT resection without adjuvant CRT. The high-risk group included: patients with T4 tumors and ≥50% BOT resection with adjuvant CRT (Table 5).

Table 5.

Classification system that stratifies swallow dysfunction at 1 year postoperatively into low, intermediate, and high risk.

This study has limitations, however, that should be considered. Due to the retrospective nature, there were occasional incomplete data regarding intraoperative findings and swallow function. Regarding swallow function evaluation, objective studies (i.e. modified barium swallow) or patient-reported outcome instruments (i.e. MDADI) were not consistently performed in this population, so the validated FOSS instrument was utilized. Due to cancer recurrence in this high-risk population as well as incomplete data, we were unable to analyze or comment on swallow function past 1 year. This limitation may have impacted our swallow outcomes, as swallow function following OPC treatment is dynamic and can change over time.¹³ Another potential limitation is that the majority of surgeries were performed by a single surgeon (61/82, 74%). As such, the data may be more internally consistent but less generalizable to other surgeons or institutions. Multiple variables also occurred with a frequency too low to make inferences on their impact on swallow function. Future studies with larger datasets should be conducted to establish this relationship.

Oncologic outcomes from surgical versus nonsurgical modalities for advanced OPC are controversial.^32–34 Survival rates are even further confounded by the increased prevalence of human papilloma virus (HPV) positivity, which is known to impart improved survival outcomes.³⁵ Due to the often unclear oncologic outcomes of surgical versus nonsurgical therapy for locally advanced OPC, treatment may ultimately be determined by surgeon or patient preference, and/or perceived quality of life factors (i.e. swallow function). Poor swallow function can be a devastating consequence of treatment for OPC and has been shown to lead to social isolation, decreased quality of life, and an overall decrease in emotional well-being.^36–38 It is critical that swallow function be an important part of any discussion regarding treatment of tumors of the oropharynx. This study provides expected risk of swallow dysfunction after primary transoral resection of advanced stage OPC, data that may ultimately help patients decide between treatment modalities.

Conclusion

This study provides predictors of swallow function at 1 year after treatment for advanced T3 and T4 OPCs removed by TOS. We observed acceptable swallowing in the majority of patients. However, BOT resection ≥50%, especially when coupled with T4 tumor stage or adjuvant CRT, was associated with poor long-term swallow outcomes.