Institutional experience with total pharyngectomy reconstruction: Exploring the role of the salivary bypass tube

There is a lack of consensus regarding the effectiveness of salivary bypass tubes during total pharyngectomy reconstruction to prevent pharyngocutaneous fistula or pharyngoesophageal stricture.


| INTRODUCTION
The total laryngectomy (TL) with total pharyngectomy (TP), generally reserved for advanced laryngeal or hypopharyngeal cancer, involves resecting a circumferential segment of the pharynx.This creates a unique challenge for the reconstructive surgeon.To accomplish the essential objectives of prompt wound healing and the re-establishment of swallowing function, a "neopharynx" can be created utilizing a tubed free tissue transfer technique (Figure 1). 1 Fasciocutaneous free flaps harvested from the radial forearm (RF), anterolateral thigh (ALT), or jejunum have been described for this purpose, yielding similar outcomes. 2,3he most common complications following pharyngeal reconstruction are pharyngocutaneous fistula (PCF) and pharyngoesophageal stricture, which can delay or prevent wound healing and return of swallow function. 1Rates of these complications range widely and depend on various risk factors.Prior radiotherapy (RT) with or without chemotherapy, hypopharyngeal primary tumor location, poor nutritional status, concurrent neck dissection (ND), and low serum hemoglobin concentration are potential risk factors for PCF. 4,5Prior PCF formation, lack of free flap reconstruction, and prior RT are potential predisposing factors for stricture formation. 1,3,6he salivary bypass tube (SBT), first introduced in 1955 by Montgomery, is a silicone device that may be placed in the neo-pharynx during reconstruction to shunt saliva past pharyngeal suture lines and stent the neopharynx open to facilitate healing (Figure 1). 7Several series have described the insertion of the SBT during pharyngeal reconstruction as a preventive measure against PCF or stricture, with some advocating its adoption into routine practice.][10][11] Regardless, the data supporting SBT use in pharyngeal reconstruction is mostly low-level and derived from series involving heterogeneous patient populations.
Herein, we report outcomes from a series of TP patients, each of whom underwent pharyngeal reconstruction with tubed fasciocutaneous RF or ALT free flaps.We compare PCF and stricture rates between those who received SBTs at the time of pharyngeal reconstruction and those who did not.We also compare our outcomes following TP to those reported in the literature.

| METHODS
This retrospective cohort study examined the records of patients who underwent total, or circumferential, pharyngectomy (TP) between 2010 and 2021 at a single academic tertiary referral center.Patients who underwent TP following prior TL and those who underwent total laryngopharyngectomy (TLP) as a single procedure were included.Patients were excluded if they had less than 6 months of follow-up with this institution, but not if they died or entered hospice care within 6 months and were followed until then.The rates of PCF and stricture were compared between those who received SBTs during initial reconstruction and those who did not.Montgomery ® SBTs (Boston Medical Products, Boston, MA) were used.Tube sizes were selected depending on the diameter of the distal cervical esophagus and the reconstructed neopharynx's diameter.The goal was to find a tube diameter that was large enough to prevent the tube from sliding, and small enough to avoid undue pressure on the anastomosis.Seven standard tube diameters are available, ranging from 8 to 20 mm.Whether patients received SBTs was at the discretion of the reconstructive surgeon, of which there were four, based on the perceived risk of PCF or stricture development in each case.PCF and stricture rates were also compared according to potential risk factors.Finally, overall complication rates, survival, speech, and swallow outcomes were examined.
Statistical analyses were performed using GraphPad Prism (version 9.3.0 for MacOS, GraphPad Software, San Diego, CA, USA, www.graphpad.com).Comparisons of PCF and stricture rates depending on multiple independent variables were analyzed using a two-sided Fisher's exact test with the cutoff for statistical significance set at p < 0.05.Kaplan-Meier survival curves were also constructed.Post hoc power analyses were conducted using G*Power open-source software. 12The Loyola University Health Sciences Center Institutional Review Board approved this study.

| RESULTS
Forty-four patients underwent TP between 2010 and 2021.Seven were lost to follow-up; one was excluded because a total esophagectomy was also performed.Thirty-six patients were included in the analysis.Mean follow-up was 26.3 months (range 0.5-120 months), with those followed for less than 6 months having died or entered hospice.Summarized demographic and surgery data are represented in Table 1.There were 24 males (67%) and 12 females with a mean age of 67.6 years at the time of TP.Thirty-three (92%) underwent TLP, and three underwent TP after previously undergoing TL.Twenty-five (69%) had prior treatment for their cancer; 11 (31%) with surgery plus RT, and 14 (39%) with RT alone.
Twenty-six (72%) defects were reconstructed using tubed RF free flaps, and 10 were reconstructed Four of the 19 patients who received SBTs were excluded from stricture rate calculations as the strictures were considered secondary to progressive or recurrent malignancy.Thus, 10 of the 15 patients considered (67%) who received SBTs developed strictures, compared with 10 of 17 (59%) of those who did not receive SBTs (RR = 1.13, 95% CI 0.64-2.0,p = 0.73).None of the PCFs that formed appeared to coincide temporally with progressive or recurrent malignancy.Eight of 19 (42%) who received SBTs developed PCF, compared with 12 of 17 (71%) if those who did not (RR = 0.47, 95% CI 0.21-0.91,p = 0.101).Neither of these differences reached statistical significance (Table 2).
Analysis of potential risk factors showed that having undergone neck dissection at the time of TP was significantly associated with a higher PCF rate ( p = 0.041).Other variables analyzed, including tumor location, prior surgery with or without RT, and type of free flap used, showed no significant association with PCF or stricture incidence (Table 2).
There were five SBT-specific complications.Four of 19 (21%) SBTs migrated out of place or rolled upon themselves, requiring removal or endoscopic retrieval from the stomach (Figure 2).One patient developed minor nasal columellar breakdown from the silk anchoring suture.
Thirteen of 15 (87%) patients with primary or secondary tracheoesophageal prostheses (TEP) gained the ability to phonate using their TEP postoperatively, while two remained aphonic.Of those who survived for at least 2 months postoperatively, 20/34 (59%) tolerated some PO intake.Five-year overall and disease-specific survivals were 41.3% and 31.7%,respectively (Figure 3).

| DISCUSSION
The TL with TP, in which a 360-degree segment of the pharynx is resected, is indicated as primary or salvage treatment for select cases of advanced malignancy involving the larynx and hypopharynx.Unfortunately, the postoperative course is often fraught with complications, the most common being PCF and stricture. 1The utility of implanting an SBT during pharyngeal reconstruction as a preventative measure against PCF or stricture formation remains a matter of debate, mainly due to the heterogeneity of the patient populations examined in available studies.Herein, we complement the existing body of literature exploring the efficacy of the SBT for these purposes.Specifically, we describe outcomes in a relatively homogeneous population of TP patients whose defects were reconstructed with tubed RF or ALT free flaps.
Approximately half (53%) of our 36 patients had SBTs placed during pharyngeal reconstruction, allowing for straightforward comparison.It should be noted, however, that SBTs were placed at the discretion of the reconstructive surgeon according to their perception of risk for PCF or stricture formation in each patient, making it likely that those in the SBT group had a higher baseline likelihood of developing these complications.As expected, PCF and stricture were the most common complications observed.

| Pharyngocutaneous fistula
4][15][16][17] Regarding our relatively high PCF rate, it should be noted that total pharyngectomy defects are significantly more challenging to reconstruct than defects that spare a portion of the pharyngeal mucosa.9][20] A previous study by our team included 59 TL patients with at least 3 cm of pharyngeal mucosa spared, and the overall PCF rate was only 23.7% after reconstruction using either a pedicled flap or free flap. 21hile we showed a lower incidence of PCF in patients with SBTs placed versus those without (42% vs. 71%), this difference did not reach statistical significance ( p = 0.107).A post hoc power analysis showed our study to be underpowered to detect a significant difference of this magnitude, with a power (1-β) of 0.37.Approximately double the sample size would be required to achieve power >0.80.Our findings align with a recent meta-analysis of five studies, demonstrating a nonsignificant trend toward lowered PCF rate in TLP patients with SBTs.Interestingly, when that group excluded studies in which less than 20% of patients received SBTs, they found a highly significant reduction in PCF rate with SBT use. 8ur analysis showed significantly higher rates of PCF in patients who had concomitant neck dissection at the time of TP (p = 0.041).This reinforces available evidence that neck dissection at the time of TP is a risk factor for PCF formation. 10Interestingly, we found no significantly higher incidence of PCF in patients who had prior RT with or without surgery.This contrasts with reports of higher PCF incidence following TL or hypopharyngectomy performed in the salvage setting. 4,6,22Our study may have been underpowered to detect such a difference.Interestingly, our study included a higher percentage of patients who received prior RT than the majority of other available studies on the topic, at 69%. 1,6,16,17 Of the patients in our group who developed PCF, 60% had prior RT.A similar finding was made by Tsou and colleagues, who found that 58.3% of 52 patients who developed PCF had prior RT.While their study showed a significant correlation between preoperative RT and PCF formation, ours did not demonstrate a statistically significant correlation. 17Most (72.2%) of our patients suffered from primary tumors of the hypopharynx.When comparing patients with hypopharyngeal primaries to all others, there was a higher incidence of PCF in the non-hypopharyngeal group (46.2% vs. 80%), though this difference was not statistically significant.This contrasts with findings in a 2015 meta-analysis that identified hypopharyngeal tumor location as an independent risk factor for PCF formation. 17A 2010 prospective randomized trial comparing RF versus ALT free flap reconstruction for 19 TLP defects found significantly higher PCF rates in the ALT flap group. 23We found no such association.

| Pharyngoesophageal stricture
Some degree of esophageal stricture developed in 62.5% of our patients, a higher percentage than the published 3% to 47% incidence following TL. 1,6,22Again, this may relate to our small sample size and the fact that all of our patients underwent circumferential pharyngectomy.360-degree pharyngeal resection appears to predispose to swallowing difficulty, with one study showing a significantly higher (45.5%) incidence of postoperative dysphagia in those with more than half of their posterior pharyngeal walls resected. 24he use of the SBT as a preventative measure against stricture remains controversial.A recent systematic review of seven studies could not draw any conclusions about the effect of the SBT on stricture risk due to the wide variation in stricture rates between studies.They found the stricture rate with SBT use after TL (but not necessarily TP) to range between three and 16%. 11We saw a higher stricture incidence in the SBT group (66.7% vs. 58.8%),though this difference did not achieve statistical significance.A post hoc power analysis revealed our study to be underpowered to detect a significant difference of this magnitude, with a power (1-β) = 0.06.Thus, whether SBT use impacts stricture risk remains to be determined.
It seems intuitive that PCF could predispose to later stricture formation, given the inflammation and scarring often accompanying salivary fistulas.A recent multicenter cohort study involving 405 salvage TL patients noted a significantly higher incidence of stricture in patients who developed PCF postoperatively. 25Others, including our group, found no such association, with our data showing identical rates of stricture formation in patients with postoperative PCF and those without (62.5%). 1 Between ALT and RF free flaps, there is no clearly superior reconstructive technique for circumferential pharyngeal defects.One study reports an association between ALT free flap reconstruction and increased stricture rate (47%).However, the comparisons were primarily made to the jejunum and latissimus dorsi free flaps, which were not used in our study. 1We found no significant association between primary tumor location and stricture rate, contrasting with multiple reports of hypopharyngeal primary tumors predisposing to stricture formation. 26,27Similarly, we found no significant association between stricture rate and concomitant neck dissection or prior RT +/À surgery.This is in keeping with several reports of preoperative RT showing no significant association with postoperative stricture rate. 26,27

| Other complications
There were no free flap failures related to vascular occlusion.There was one perioperative (within 1 month) death in a patient who experienced bleeding on postoperative day (POD) 14 and was taken to surgery for wound exploration, revealing a salivary fistula and flap dehiscence from the esophagus.The patient was transitioned to hospice care and expired on POD 20.Five of 19 patients (26.3%) had minor SBT-related complications, including two SBTs that migrated out of place, requiring endoscopic retrieval from the stomach (Figure 2), and two that rolled upon themselves and required removal.One patient developed minor nasal columellar breakdown from the silk anchoring stitch.None of these complications resulted in significant morbidity.SBT-related complications appear to be rare, with another study reporting displaced tubes in just three of 69 patients with no significant morbidity. 10

| Survival, speech, and swallowing outcomes
Our group demonstrated 5-year disease-specific and allcause survival of 41.3% and 31.7%,respectively (Figure 3).This aligns with quoted disease-specific and overall 5-year survival rates of 25%-56% and 11%-47% after TLP, respectively. 28,29Of 15 patients with primary or secondary tracheoesophageal punctures (TEP), 13 (87%), or 36.1% of our patients, achieved esophageal speech.As of their most recent follow-up visits, 20 (55.6%) tolerated peroral (PO) intake, and nine patients did not require supplemental tube feeds.These numbers are consistent with other studies, including one in which 80% of TEP recipients had intelligible speech following TLP, and 30% could swallow. 1

| CONCLUSION
Survival figures, speech, and swallow outcomes were acceptable in 36 patients who underwent TP and reconstruction with tubed ALT or RF free flaps.The role of the SBT in TP reconstruction remains controversial.We show a nonsignificant reduction in PCF incidence with SBT use, adding to a largely inconclusive body of evidence.The utility of the SBT for stricture prevention appears minimal.Larger randomized and controlled studies are necessary to determine whether the SBT has a role in TP reconstruction.

F
I G U R E 1 Neo-pharynx fashioned using a tubed radial forearm free flap technique.Flap after harvest (left) and during inset with salivary bypass tube in place (right).[Color figure can be viewed at wileyonlinelibrary.com] using tubed ALT free flaps.Nineteen (53%) had SBTs placed during reconstruction, and 17 did not.The SBTs were left in place for an average of 27.2 days (range: 5-124 days).

F
I G U R E 3 Kaplan-Meier curves demonstrating disease-specific (A) and all-cause mortality (B).
Demographic and surgery data.
T A B L E 2 Incidence of fistula (PCF) and stricture according to various risk factors.
Abbreviations: RT, radiation therapy; SBT, salivary bypass tube.F I G U R E 2 Endoscopic photo of displaced salivary bypass tube in gastric body.[Color figure can be viewed at wileyonlinelibrary.com]