Neopharyngeal Stricture following Laryngectomy

Abstract

Stricture formation is a serious complication following pharyngeal reconstruction. These strictures can be life-threatening and can severely impact quality of life. In this article, the existing literature on surgical risk factors linked to neopharyngeal stricture formation is reviewed. Intraoperative preventative measures reconstructive surgeons should consider are also discussed. Finally, this article will describe the evaluation and management of pharyngoesophageal strictures, including the challenges and options when dealing with refractory strictures.

Reconstruction of the pharynx following laryngectomy provides patients with a means to speak and swallow; however, complications can arise. One of the most common is neopharyngeal stricture. Strictures often present themselves in the context of oropharyngeal dysphagia (OD). OD is a common symptom following laryngectomy, affecting anywhere from 10% to 60% of patients. ¹ Presenting symptoms include feeding intolerance, nasopharyngeal regurgitation, globus sensation, and prolonged mealtimes. Dysphagia usually begins with solids, but may progress to liquids if left unmanaged. ^{2 3} Tools to evaluate swallowing function in these patients include flexible endoscopic evaluation of swallow, manometry, and, most frequently, video fluoroscopic swallowing study. ¹ Strictures have been reported to account for up to one-third of patients with OD following laryngectomy. ⁴ These strictures are defined as a static pharyngeal narrowing affecting either swallowing or speech. ⁵ It has been reported that 82% of strictures form within the first year following surgery. ⁶

Depending on the extent of surgery, strictures may form in the neopharynx or proximal esophagus. These strictures not only affect quality of life but also can lead to prolonged gastrostomy tube dependency and severe malnutrition. Pharyngeal strictures may also impair patients' ability to achieve fluency with tracheoesophageal speech. ⁷ These strictures should be identified and managed as early as possible through endoscopic dilation and, in some cases, further surgery through pedicled or free-tissue transfer. Known risk factors for stricture formation include closure technique, pharyngocutaneous fistula (PCF) formation, and concomitant chemoradiotherapy (CCRT). ^{8 9 10}

In this article, we will describe the relationship of these risk factors and management of pharyngeal structure after total laryngectomy. While strictures may form in all cases of pharyngeal reconstruction, the current literature primarily discusses their incidence following laryngectomy. Issues with incomplete or recurrent cricopharyngeal hyperfunction are beyond the scope of this article.

Closure Technique

Closure of pharyngeal defects can be accomplished through a variety of techniques. For partial pharyngeal defects, surgeons may elect to close primarily with no flap, placing a flap over the primary closure (“onlay flap”), or use pedicled or free flaps through “patch” closure. For circumferential pharyngeal defects, a multitude of free flaps have been studied to be effective, ranging from fasciocutaneous to enteric for defects of the hypopharynx also involving the cervical esophagus. No technique is resistant to stricture, as all have shown stricture formations with varying rates.

Primary Closure

Under appropriate conditions, primary closure without the use of a flap may be used following partial laryngectomy. Generally, at least 1.5 cm of relaxed or 2.5 cm of stretched mucosa is sufficient to close primarily. ¹¹ Stricture rates for these patients have been shown to range anywhere from 2.3% to 43.9%. ^{6 12 13 14} Some data suggest primary closure affords the patient superior swallowing outcomes when compared with pedicled or free flaps. ¹⁵ In regard to stricture formation, primary closure has been linked to higher rates when compared with flap closure. ^{1 6 16} One explanation for this observation is that a primarily closed pharynx will result in a reduced lumen diameter compared with patch or tubed reconstruction, potentially increasing the likelihood of stenosis or stricture. ¹⁷ Although higher rates of stricture formation have been reported, strictures in these patients were more likely to be successfully managed by a single balloon dilation compared with flap closure. ⁶ Further studies have been conducted to examine whether orientation of pharyngeal closure affects stricture rates. While it was found that “t-type” closure led to decreased risk of PCF when compared with vertical closure, there has been no observed difference in stricture rates between the two closure patterns. ¹²

In some cases, surgeons may elect to place a myofascial flap over the primary pharyngeal closure. These onlay flaps have been studied mainly in the context of salvage laryngectomy (SL), and have been shown to provide significantly lower ( p  < 0.05) PCF rates when compared with primary or patch closure. ^{17 18 19} Similar to primary closure, use of an onlay flap has been observed to have significantly higher rates of dysphagia and stricture formation when compared with patch closure. ^{20 21} There are little data examining differences in stricture rates in patients closed primarily versus those with onlay flaps.

Patch Reconstruction

Without sufficient residual mucosa, partial pharyngeal defects may require patch reconstruction. This can be accomplished through the use of pedicled flaps, classically the pectoralis major myocutaneous (PMMC) flap, or through the use of a variety of fasciocutaneous free flaps (FCFFs). As stated earlier, patients undergoing flap closure have been shown to be at lower risk of stricture compared with primary closure. Furthermore, partial pharyngeal defects closed with patches are less likely to form stricture compared with circumferential defects closed with tubed flaps. ⁶

A wide variety of pedicled flaps have been described for pharyngeal reconstruction, including PMMC, latissimus dorsi myocutaneous (LDMC) flap, thoracodorsal artery perforator, sternocleidomastoid myocutaneous (SCM) flap, and supraclavicular flap. ^{21 22 23 24 25 26 27 28 29} Among these, the PMMC is the most widely used and is the only flap for which stricture data are readily available. Reviews have shown this flap to carry a mean stricture rate ranging from 2 to 27%. ²² Though data on stricture rates are lacking, PCF rates for other pedicled flaps have been reported to be lower than that of PMMC. Compared with PMMC, which has shown to carry a mean fistula risk of approximately 54%, PCF rates for LDMC and supraclavicular flaps have been reported to be as low as 6 and 16%, respectively. ^{24 29}

Fasciocutaneous free tissue reconstruction of partial pharyngeal defects has been proven to be effective through the use of radial forearm free flaps (RFFFs), anterolateral thigh free flaps (ALTFFs), and supraclavicular artery island flaps (SCAIFs). ²¹ SCAIF is not as commonly utilized but has not been shown to significantly impact PCF nor stricture rates when compared with RFFF and ALTFF. Therefore, SCAIF is a viable and valuable option for patch reconstruction. ²¹

There are some obvious advantages to using FCFF versus pedicled flaps for patch reconstruction. Pedicled flaps such as PMMC tend to be bulkier and less pliable than FCFF, making it more difficult to tailor to partial or near-circumferential defects. Furthermore, FCFF reconstruction has been shown to produce superior swallowing outcomes compared with pedicled flaps. ³⁰ Modifications to patch FCFF to reduce stricture rates even further have also been described. One study describes fashioning the distal portion of the flap into an “arrowhead” while also creating a 1.5-cm slit at the level of the anterior wall of the distal anastomosis site ( Fig. 1 ). ³¹ This was done to reduce the likelihood of circumferential stricture formation, and yielded promising results with a stricture rate of 1.8%. However, it is still uncertain whether type of flap used for patch reconstruction is a predictor for stricture formation. Although one study has pointed out the patients with PMMC tend to need more serial dilations due to stricture compared with those receiving FCFF, 25% and 9%, respectively, a significant difference has yet to be reported. ³¹

Fig. 1.

“Arrowhead” flap for patch reconstruction of pharyngeal defect. ( A ) Radial forearm free flap fashioned into an “arrowhead” shape ( white asterisk ) with external paddle ( black asterisk ). ( B ) Insertion and suturing of “arrowhead” into 1.5-cm slit created in the anterior wall of the pharynx ( white asterisk ).

Tubed Reconstruction

In the event of a circumferential pharyngeal defect, as in cases of total laryngectomy, a “tubed” flap reconstruction is indicated. Patients requiring reconstruction of circumferential defects are more susceptible to stricture formation compared with those closed primarily and those closed with a patch. ^{6 32} An explanation for this finding is that the circumferentially inset FCFF is not able to stretch to the degree that natural mucosa is able to and it is more likely to scar and lead to narrowing of the lumen. For this reason, efforts have been made to widen the anastomosis and interrupt the circular suture line. One of these interventions is spatulation with interdigitation of the distal anastomosis site. This practice involves making longitudinal incisions at 120-degree intervals at the distal end of the tubed flap and distal anastomosis site. The ends are then interposed forming a zig-zag pattern, interrupting the otherwise circular suture line. ³³ Other methods to reduce stricture include utilizing larger flap sizes and triangular or trapezoidal extension of the flaps. ^{33 34 35 36 37 38 39 40 41 42 43}

When comparing differences in stricture rates among flaps used for tubed reconstruction, meta-analyses and comparative studies have found enteric flaps to have lower stricture rates than FCFF. ^{44 45 46 47} Enteric flaps, most often in the form of free jejunal transfer, are thought to have lower incidence of strictures due to the resulting mucosa–mucosa anastomosis as opposed to the mucosa–cutaneous anastomosis produced with FCFF. ⁴⁷ Other studies went on to examine PCF and stricture incidence in jejunal flaps after being reinforced with a pedicled flap, such as PMMC or SCM. These studies showed that this flap reinforcement reduces fistula formation, and one study found SCM reinforcement reduces stricture formation at anastomotic sites. ^{48 49} Enteric tissue transfer has been subject to some criticism, including resulting speech having a “wet” quality, worse short-term swallowing outcomes, and high donor-site morbidity. ^{47 50 51 52} However, recent meta-analyses have found no significant differences in donor-site morbidity or speech and swallowing rates when comparing jejunal flaps and FCFF. ^{44 46}

The current data on differences in stricture rates between ALTFF and RFF for circumferential defects have demonstrated conflicting results. Randomized prospective trials have shown a significant reduction in stricture incidence in RFFF compared with ALTFF. ⁵³ Reasoning for this difference is that ALTFF introduces more bulk and may reduce lumen diameter by sheer mass. However, other studies have shown higher, though not significant, rates of stricture in RFFF compared with ALTFF. ^{47 54} Furthermore, ALTFF has been shown to significantly reduce formation of PCF compared with RFFF. ³² This is attributed to the inclusion of the fascia lata for reinforcement of the suture line. ⁵⁵ However, meta-analyses have shown no significant differences in stricture formation between the two flaps. ^{44 46 47} Pedicled flaps, such as PMMC, have also been used to repair circumferential defects as well, but have fallen out of favor due to reported high rates of stricture, longer fistula duration, and poorer speech and swallowing outcomes. ^{22 30}

Primary versus Salvage Laryngectomy

Researchers have also attempted to identify differences in stricture rates between patients undergoing primary laryngectomy (PL) and those requiring SL for persistent or recurrent disease after failure of CCRT. Radiation therapy (RT) has already been proven to be a risk factor for strictures, possibly putting patients requiring SL at increased risk for stricture. ⁵⁶ CCRT/RT induces fibrosis in the neck, which can be scored using the “woody hardness” classification. ⁵⁷ Using this classification, a significant association was found between increasing degree of fibrosis and time to stricture development. ⁵⁸ Strictures seem to be more common and severe depending on the severity of neck fibrosis/woodiness.

However, research has shown that patients needing SL are two times more likely to be reconstructed with a flap, while those undergoing PL are three times more likely to be closed primarily. ⁶ Primary closure has been shown to be a risk factor for stricture, ^{1 6 20} which might correlate with higher stricture rates after PL. As of now, however, there has been no observed significant difference in stricture rates or increased need for dilations between PL and SL.

Pharyngocutaneous Fistula

PCF is an extensively studied early complication following laryngectomy. Various risk factors for the formation of PCF have been identified including SL, chemoradiation, low albumin, and pharyngeal reconstruction method. ^{16 59} PCF has also been described as a risk factor for stricture development and an increased need for dilation. ^{8 9} If left untreated, long-standing fistulas result in granulation tissue and healing by secondary intention. This can progress to scar contracture and promote stricture formation. ⁶⁰ Some researchers have observed that it is not uncommon for patients with a history of PCF to develop treatment-refractory strictures well after fistula has resolved. ⁶¹ It is likely that the inflammation for the PCF leads to increased scar contracture in the neopharynx and, thus, stricture formation. It is then reasonable to conclude that prevention of PCF could also be protective against stricture formation.

Salivary Bypass Tubes

Salivary bypass tubes (SBTs) in combination with pharyngeal reconstruction have been shown to reduce incidence of PCF and stricture formation. ^{62 63 64 65} When examining risk factors for PCF, use of SBT was associated with a lower risk of PCF by multivariate analysis. ⁶⁴ Additionally, patients in this study receiving SBT were found to have significantly lower rates of stricture formation compared with those without SBT. SBTs have also been shown to reduce hospital stay and time to fistula closure. ^{65 66} It is thought that SBTs help stent reconstruction during surgery, decrease salivary exposure of the anastomotic suture line, and therefore aid in the healing of the neopharynx. ⁶³ Another study placed an SBT directly after endoscopic dilation of pharyngeal stricture and removed it after 50 days. They found that only 13.3% of these patients had a recurrence of their stricture after 6-month follow-up, lower than a previously reported 25% recurrence rate in a randomized controlled trial of endoscopic dilation of strictures. ^{67 68}

A meta-analysis of SBT use in laryngectomy found that time to removal of SBT ranged from 10 to 60 days, with a mean of 21.8 days. ⁶⁹ There was marked heterogeneity in the studies included in analysis, especially when those where less than 20% of the patients were given SBT during surgery were included. When these were excluded, statistical significance was reached in regard to a lower risk of PCF formation with the use of SBT. ⁶⁹

SBTs have been employed in all forms of reconstruction and defect types; however, analysis on data regarding direct comparisons among different reconstruction techniques is lacking. Although not reaching statistical significance, SBTs have also been shown to produce low stricture rates when used in conjunction with PMMC, RFFF, and ALTFF. ^{34 62 70}

Management of Stricture

Etiologies and Characteristics

The literature on pharyngoesophageal stricture from the endoscopist's perspective is extensive and provides valuable lessons that are applicable to the reconstructive surgeon's patient population. In all cases, appropriate management of strictures begins with identification of etiology. Generally, strictures may be categorized as either benign or malignant. Malignant strictures caused by tumors such as adenocarcinomas, squamous cell carcinomas, and metastasis are managed through curative surgery. Benign strictures, such as anastomotic pharyngeal strictures described in this review, have historically been managed by medication, surgical reconstruction for refractory strictures, and most commonly through endoscopic dilation. ^{71 72} Other etiologies of benign strictures include corrosive caustic ingestion (e.g., household cleaning products), eosinophilic esophagitis, drug-induced esophagitis (e.g., nonsteroidal anti-inflammatories, tetracyclines), infectious esophagitis, radiation or chemotherapy injury, thermal injury, and iatrogenic strictures after endoscopic therapy. ⁷²

Through contrast studies or endoscopy, benign strictures may be further defined as simple or complex. Simple strictures are straight and short, usually less than 2 cm, and easily allow passage of an endoscope. Complex strictures are longer than 2 cm and have an uneven surface, tortuous margins, and a narrow diameter. ⁷³ The majority of patients are evaluated with endoscopy when stricture is suspected. Once presence of stricture is confirmed, it is then biopsied to determine etiology and guide management. ⁷³

Treatment

The ultimate goal of stricture treatment is to improve dysphagia by re-establishing luminal patency of the pharynx or esophagus. Dysphagia of solids is said to occur with esophageal lumens less than 12 mm, and, in general, dilation of at least 16 mm and up to 20 mm is common practice. ^{71 74 75} Endoscopic dilation is the most common treatment strategy for benign strictures. These dilations may be performed under general anesthesia, sedation, or in the unsedated patient. ⁷⁶ A national survey of endoscopists' experience showed that cardiopulmonary events secondary to conscious sedation account for most complications during endoscopy. ⁷⁷

Endoscopic dilation is done by means of either a bougie or balloon dilator. Bougie dilators include Maloney, Savary-Gilliard (SG), and less commonly Hurst and Eder dilators. ⁷⁸ Maloney dilators are used without a guidewire normally for distal, simple strictures with lumens greater than 10 mm. ⁷⁹ However, use of Maloney dilators is associated with increased risk of perforation for complex strictures and those with narrow lumens. ⁸⁰ In these situations, a wire-guided SG bougie is preferred. SG bougies have been shown to be more cost-effective and have been associated with greater improvement of stricture diameter compared with balloon dilators. ^{3 74 81} A proposed advantage of balloon dilators is the application of force in a radial fashion as opposed to the longitudinal forces applied by bougies. There has been speculation if this increase in longitudinal forces make bougies more likely to result in increased bleeding or perforation. ⁸² However, SG bougies have provided very low complication rates and there has been no significant difference observed when compared with balloon dilation. ^{74 81}

While not providing a reduction in complications, balloon dilation is more easily applied in cases of tortuous complex strictures than bouginage. ^{80 83 84} Technique is described in Fig. 2 . Regardless of type of reconstruction, endoscopic balloon dilation of neopharyngeal strictures has been proven to be safe and effective. Balloon dilation can provide symptomatic relief, improve dietary outcomes, and reduce pharyngeal resistance during swallowing. ^{1 81 85 86}

Fig. 2.

Balloon dilation of neopharyngeal stricture. ( A ) Stricture is visualized by endoscopy. ( B ) Deflated balloon is passed on guidewire to level of stricture by endoscopic guidance. ( C ) Balloon is inflated to dilate stricture. ( D ) Visualization of lumen postdilation.

In some instances, stenosis due to stricture can be so severe that there is complete or near-complete obliteration of the pharyngoesophageal lumen. Endoscopic dilation with the use of bougies or balloons is impossible in these cases. A technique known as combined antegrade and retrograde dilation (CARD) has been used for complete or near-complete stenosis. ^{87 88} This two-person procedure involves proximal approach using a direct laryngoscope or rigid esophagoscope to visualize the stricture, followed by distal approach through the patient's gastrostomy tube site. After blunt dissection or sharp incision from the proximal end, a guidewire is passed from the distal side of the stricture, grasped by the operator on the proximal side, and then brought through the mouth. A wired-guided SG bougie can then passed through the created lumen to dilate the stricture. ⁸⁸ Some surgeons may elect to keep a nasogastric tube in place to serve as a temporary stent. Complication rates for this procedure are high and studies have shown that only a minority of the patients achieve sufficient oral intake to discontinue tube feeds altogether. However, CARD has been shown to significantly improve swallow function, and therefore should be considered in patients before committing them to permanent tube-feed dependence. ^{87 88}

Refractory Strictures

Unfortunately, anastomotic strictures often recur after treatment, reportedly as high as 25% in 6 months and 50% after a year, and may become refractory after multiple dilations. ⁶⁸ These refractory strictures have been defined as an anatomic restriction because of cicatricial luminal compromise or fibrosis causing dysphagia in absence of inflammation. ^{61 89} These are diagnosed clinically when luminal patency of at least 14 mm is not attained after five or more dilations at intervals of 2 to 4 weeks. ⁸⁹ One identified risk factor for development of refractory strictures was found to be use of fluoroscopy during dilation. This may be a selection bias due to use of contrast guidance in anticipation of complex stricture, which are more prone to be refractory. ⁹⁰ However, prior neoadjuvant chemotherapy was found to significantly reduce the risk of refractory strictures compared with no chemotherapy. This may be due to reduced postoperative fibrosis, leading to reduced luminal structuring. ⁹⁰ Proposed treatments for these refractory strictures have included topical or intralesional mitomycin C, esophageal stents, home self-dilations using Maloney bouginage, and even through regional or free flap tissue reconstruction. ^{60 61 81 90 91}

Self-expanding metal stents (SEMSs) are normally reserved for palliative treatment in the case of advanced esophageal cancers causing malignant stricture. However, SEMS have also been employed for the treatment of benign refractory strictures. ^{61 72} Conventional endoscopic stricture dilation with bougie or balloon is performed, followed by passing of the SEMS over the same guidewire. Under endoscopic guidance, the SEMS is passed to the site of the stricture and deployed, and left for 4 to 8 weeks before removal. ⁹² These stents may be used for patients where repeated dilations are undesirable. A meta-analysis demonstrated that stents are effective in approximately 40% of patients. However, the high migration and adverse event rate, reported at 28.6% and 20.6%, respectively, has discouraged their use in benign strictures. ⁹³

Two recent publications aimed to report experiences using free-flap reconstruction for strictures following laryngectomy. One of these studies opted to FCFF reconstruction for these “end-stage” strictures and has reported short-term success. ⁹⁴ At follow-up (median time of 14.1 months), four of five patients were tolerating a full oral diet, with one patient was dying during the study period due to disease recurrence. Although two of these four patients did require further dilations, none of the patients required tube feeding. The other study elected for right colonic transposition in two patients with refractory stricture after FCFF reconstruction. ⁹⁵ Previous studies have reported lower stricture rates for enteric flaps compared with FCFF, such as the jejunal free flap. ⁴⁵ Though both patients had previously been dependent on gastrostomy tube for intake, both patients tolerated a full oral diet at 17 (mean) months' follow-up.

Conclusion

Pharyngeal stricture formation is multifactorial and can be challenging obstacle for patients following pharyngeal reconstruction. Its association with PCF should urge surgeons to take measures to curtail fistula formation, possibly through the use of onlay flaps when defects are closed primarily and through implementation of SBT. These efforts have been shown to not only reduce fistula formation but also decrease the need for serial dilations and incidence of refractory strictures. Furthermore, free and regional tissue transfers for partial pharyngeal defects have resulted in lower stricture rates and superior swallowing outcomes compared with primary closure with or without onlay. When possible, surgeons should consider using FCFF before pedicled flaps for partial pharyngeal defects due to ease of tailoring, evidenced lower dilation rates, and superior swallowing outcomes. For circumferential defects, current literature suggests enteric flaps reduce the incidence of stricture when compared with FCFF.

Although a significant difference is yet to be observed, future studies may yield a clearer relationship with stricture incidence and primary surgery versus organ preservation. Though jejunal flaps have been thought to yield poorer functional outcomes compared with FCFF, recent data showing similar outcomes to FCFF may warrant further investigation and practice with this technique. In the event where stricture does occur, reconstructive surgeons should be aware of steps to evaluate and understand the appropriate tools to treat these strictures.