Introduction
Laryngeal cancer remains a significant health concern, particularly in countries with high smoking rates. Squamous cell carcinomas (SCC) constitute over 90 per cent of laryngeal cancers, and these are often diagnosed in elderly patients. Total laryngectomy is a common surgical approach for advanced-stage laryngeal cancers, especially in regions with delayed hospital admissions.
Since total laryngectomy is a frequent procedure, post-operative complications are a crucial consideration. Pharyngocutaneous fistula is the most common complication reported in the literature, with an incidence range of 5–58 per cent.Reference Casasayas, Sansa, García-Lorenzo, López, Orús and Peláez1 Typically, pharyngocutaneous fistula manifests during the first post-operative week after oral feeding commences.Reference Do, Chung, Chang, Kim and Rho2 This complication significantly disrupts the recovery process by delaying oral feeding, adjuvant radiotherapy (RT) and chemotherapy treatments, ultimately increasing both morbidity and mortality rates.Reference Casasayas, Sansa, García-Lorenzo, López, Orús and Peláez1,Reference Aydin, Taskin, Orhan, Altas, Ege and Yucebas3
Several factors are thought to influence pharyngocutaneous fistula development, including pre-operative low blood albumin and haemoglobin levels, comorbid diseases (e.g. diabetes mellitus and congestive heart failure), neck dissection type, tumour location and stage, and positive surgical margins post-surgery.Reference Casasayas, Sansa, García-Lorenzo, López, Orús and Peláez1,Reference Lansaat, van der Noort, Bernard, Eerenstein, Plaat and Langeveld4,Reference Wang, Xun, Wang, Lu, Yu and Guan5 Recent studies have also explored the role of pharyngeal reconstruction techniques in pharyngocutaneous fistula formation. These studies compared linear and T-shaped closures for pharyngeal closure planes, and suture techniques such as Cushing, Zipper, simple interrupted sutures, and staplers.Reference Dedivitis, Aires, Pfuetzenreiter, Castro and Guimarães6–Reference Chotipanich and Wongmanee9 However, there is a lack of consensus on the definitive risk factors for pharyngocutaneous fistula development, as different studies have identified varying factors.
This retrospective study aimed to investigate the risk factors hypothesised to be associated with pharyngocutaneous fistula development.
Materials and Methods
This study included patients who underwent total laryngectomy and bilateral neck dissection for SCC of the larynx at the Otorhinolaryngology Department of Haseki Training and Research Hospital between 2009 and 2021. We excluded patients with involvement of the hypopharynx, total or partial pharyngectomy, base of the tongue resection, large pharyngeal defects requiring free or pedicle flap reconstruction, salvage laryngectomy following RT failure, or partial laryngectomy.
Patient characteristics, medical history, preoperative examinations, blood test results, radiological imaging findings, surgery notes, follow-up data, and pathology reports were retrieved from the hospital information management system. Cut-off values were used to define specific conditions: hypoalbuminemia (albumin levels below 3.5 g/dL), anaemia (haemoglobin levels below 12.5 g/dL), and hypothyroidism (thyroid-stimulating hormone (TSH) levels above 4 mIU/L).Reference Boscolo-Rizzo, De Cillis, Marchiori, Carpenè and Da Mosto10–Reference Brabant, Beck-Peccoz, Jarzab, Laurberg, Orgiazzi and Szabolcs12 Patients were categorised into groups based on an age limit of 60 years.Reference Boscolo-Rizzo, De Cillis, Marchiori, Carpenè and Da Mosto10 Close surgical margins were considered as ≤ 5 mm.Reference Alicandri-Ciufelli, Bonali, Piccinini, Marra, Ghidini and Cunsolo13 Alcohol consumption was defined as consuming at least three glasses of alcohol daily.Reference Lemaire, Schultz, Vergez, Debry, Sarini and Vairel14
Surgical technique
All patients underwent laryngectomy following neck dissection. Two pharyngeal reconstruction techniques were employed: the T-shaped simple interrupted sutures and the vertical modified Cushing suture. Polyglactin 3/0 suture material with a round needle was used consistently across all surgeries. Notably, the larynx removal and pharyngeal reconstruction were performed by specialists with equivalent surgical experience.
Modified Cushing suturing method
The Modified Cushing suture is a continuous suture technique that, like the standard Cushing method, avoids penetrating the mucosa.Reference Deniz, Ciftci and Gultekin8 This preserves the mucosal blood supply. This technique incorporates three key modifications. The first modification resembles the zipper suture, due to its oblique transition.Reference Haksever, Akduman, Aslan, Solmaz and Ozmen7 As illustrated in Figure 1, the needle enters a point (a) farther from the suture line (1 mm) and exits (a’) closer to the suture line (0.5 mm) in one manoeuvre, without penetrating the mucosa. Importantly, the entry and exit points are not aligned vertically or horizontally. This oblique passage, with the assistance of another surgeon, spontaneously inverts the defect edge. For the second modification, on the opposite side of the defect, the entry point is located 1 mm behind the first exit point (a’) (labelled b in Fig. 1). This ensures each suture starts 1 mm behind the corresponding side, providing additional reinforcement for a watertight closure. The final modification involves knotting the suture on itself after four stitches to secure the suture line. This is crucial because a single defect in continuous sutures could tear open the entire neopharynx. This process is repeated until the defect is entirely closed (Fig. 1).
Figure 1. Modified Cushing suturing method. (a) Sutures do not disturb mucosal blood supply by not penetrating mucosa. (b) Closer look at top view. The needle enters from A (1 mm away from the edge) and exits from A′ (0.5 mm away from the edge) with an oblique manoeuvre. On the corresponding side, the needle enters from B (1 mm behind of A′) and exits from B′. Knots can be seen on D′ and H′.
Statistical method
Statistical analysis was performed using SPSS 27.0 for Windows (SPSS Inc., Chicago, IL, USA). Descriptive statistics were employed to summarise the data. Categorical variables, such as patient demographics or medical history elements, were presented as frequencies and percentages. Numerical variables, such as age or blood test results, were presented using mean, standard deviation (SD), minimum, maximum, and median. The Kolmogorov–Smirnov test was used to assess whether the distribution of numerical variables was normal. The independent samples t-test and Mann–Whitney U test were used to compare quantitative data. The chi-square test was used to compare the qualitative data. The alpha significance level was set at p < 0.05.
This retrospective study was approved by the Clinical Research Ethics Committee of the University of Health Sciences, Haseki Training and Research Hospital (05.05.2021; decision no: 15-2021).
Results
This study included 164 patients, of whom 10 (6.1 per cent) were female and 154 (93.9 per cent) were male. The mean age was 60.3 ± 10 years (range 36–89 years). Over two-thirds (111, 67.7 per cent) of the patients were younger than 60 years. Comorbidities included diabetes mellitus in 29 patients (17.7 per cent) and hypertension in 38 patients (23.2 per cent). Smoking history was present in 107 patients (65.2 per cent).
Most tumours (145, 88.4 per cent) were transglottic, and 115 patients (70.1 per cent) had T4 tumours. Most patients (145, 88.4 per cent) underwent bilateral lateral neck dissection, while the remaining patients (19, 11.6 per cent) underwent ipsilateral radical neck dissection and contralateral lateral neck dissection. Pre-operative tracheostomy was performed in 46 patients (28 per cent).
Pre-operative blood tests revealed hypoalbuminemia in 50 patients (30.5 per cent) and anaemia in 21 patients (12.8 per cent). The hyoid bone was removed from 149 patients (90.9 per cent). Surgical margins were negative in 127 patients (77.4 per cent).
For pharyngeal reconstruction, T-shaped simple interrupted sutures were used in 110 patients (67.1 per cent) and the vertical modified Cushing suture in 54 patients (32.9 per cent). Fifty-two patients (31.7 per cent) developed pharyngocutaneous fistula. Patient demographics are presented in Table 1. Evaluation of various factors including age, sex, smoking and alcohol consumption, comorbidities, tumour location, stage and differentiation, pre-operative albumin, TSH, and haemoglobin levels, pre-operative tracheotomy, and surgical margins revealed no statistically significant differences between the pharyngocutaneous fistula (+) and pharyngocutaneous fistula (−) groups. Mean pre-operative haemoglobin levels were similar between groups (p = 0.324).
Table 1. Demographic characteristics; COPD = chronic obstructive pulmonary disease; Hgb = haemoglobin; SD = standard deviation; TSH = thyroid-stimulating hormone
However, the type of neck dissection and pharyngeal reconstruction technique showed significant associations with pharyngocutaneous fistula development. In the radical neck dissection group, 12 of 19 patients developed pharyngocutaneous fistula, compared to 40 of 145 in the lateral neck dissection group. Similarly, the pharyngocutaneous fistula rate was higher in the T-shaped simple interrupted suture group (44 of 110 patients) compared to the modified Cushing suture group (8 of 54 patients; Fig. 2). These differences were statistically significant (p = 0.002 and p = 0.001, respectively) (Table 2). Univariate regression analysis identified radical neck dissection (odds ratio: 4.50, 95 per cent confidence interval [CI] 1.65–12.24, p = 0.003) and pharyngeal reconstruction with simple interrupted suture (odds ratio: 3.83, 95 per cent CI 1.65–17.00, p = 0.002) as potential risk factors for pharyngocutaneous fistula. Multivariate analysis confirmed that pharyngeal reconstruction with simple interrupted sutures (odds ratio: 3.12, 95 per cent CI 1.31–17.00, p = 0.010) and radical neck dissection (odds ratio: 3.16, 95 per cent CI 1.13–8.82, p = 0.028) were independent predictors of pharyngocutaneous fistula formation (Table 3).
Figure 2. (a) Number of PCF (-) and PCF (+) patients with lateral and radical neck dissection. (b) Number of PCF (-) and PCF (+) patients with simple interrupted suture and modified Cushing suture. PCF = pharyngocutaneous fistula.
Table 2. Comparison of PCF (−) and PCF (+) groups; PCF = pharyngocutaneous fistula; Hgb = haemoglobin; SD = standard deviation; TSH = thyroid-stimulating hormone
t Independent samples t test; m Mann–Whitney U test; X2 chi-square test (Fisher's exact test)
Table 3. Regression analysis; CI = confidence interval; OR = odds ratio
Discussion
This study investigated risk factors for pharyngocutaneous fistula, the most common complication following total laryngectomy. While the reported literature cites pharyngocutaneous fistula incidence of 5–58 per cent,Reference Casasayas, Sansa, García-Lorenzo, López, Orús and Peláez1 our study observed a rate of 31.7 per cent.
The association between diabetes mellitus and pharyngocutaneous fistula development is controversial. Studies by Mattioli et al. and Cavalot et al. suggest hyperglycaemia due to insulin deficiency or resistance following surgery can impair wound healing, increasing pharyngocutaneous fistula risk.Reference Boscolo-Rizzo, De Cillis, Marchiori, Carpenè and Da Mosto10,Reference Mattioli, Bettini, Molteni, Piccinini, Valoriani and Gabriele15,Reference Cavalot, Gervasio, Nazionale, Albera, Bussi and Staffieri16 However, other studies have not found a link, possibly due to stricter blood-sugar control before and after surgery in these patients.Reference Do, Chung, Chang, Kim and Rho2,Reference Deniz, Ciftci and Gultekin8,Reference Lemaire, Schultz, Vergez, Debry, Sarini and Vairel14
Redaelli de Zinis et al. found a connection between congestive heart failure and pharyngocutaneous fistula. They also found that congestive heart failure is related to hypertension.Reference de Zinis, Ferrari, Tomenzoli, Premoli, Parrinello and Nicolai17 Additionally, Galli et al. reported a correlation between fistula formation and cardiomyopathy, possibly due to microangiosclerosis and impaired wound healing, although the exact mechanism remains unclear.Reference Galli, De Corso, Volante, Almadori and Paludetti18 Consistent with Aydin et al.'s findings, our study did not identify a significant association between comorbid diseases and pharyngocutaneous fistula development.Reference Aydin, Taskin, Orhan, Altas, Ege and Yucebas3
Some studies have suggested that low pre-operative or post-operative haemoglobin levels are risk factors for pharyngocutaneous fistula.Reference Paydarfar and Birkmeyer11,Reference Cavalot, Gervasio, Nazionale, Albera, Bussi and Staffieri16,Reference de Zinis, Ferrari, Tomenzoli, Premoli, Parrinello and Nicolai17–Reference Morton, Fielder and Dorman21 These two parameters reflect the patient's nutritional status and can potentially affect wound healing. However, similar to the findings in other studies, our study did not observe a correlation between pre-operative anaemia and pharyngocutaneous fistula.Reference Do, Chung, Chang, Kim and Rho2,Reference Aydin, Taskin, Orhan, Altas, Ege and Yucebas3,Reference Deniz, Ciftci and Gultekin8,Reference Qureshi, Chaturvedi, Pai, Chaukar, Deshpande and Pathak22
Different studies use varying haemoglobin levels as anaemia (e.g., 12 g/dL or 9 g/dL), potentially affecting the association with pharyngocutaneous fistula development. In addition, studies employ different cut-off levels (e.g., 12 g/dL, 12.2 g/dL).Reference Boscolo-Rizzo, De Cillis, Marchiori, Carpenè and Da Mosto10,Reference Erdag, Arslanoglu, Onal, Songu and Tuylu20 Based on these observations, we propose two key points for future research: establishing a standardised haemoglobin cut-off level for defining anaemia in pharyngocutaneous fistula studies, and reporting the mean haemoglobin values for patients with and without pharyngocutaneous fistula.
Regarding tumour stage as a risk factor, some studies suggest that extensive mucosal resections associated with higher T stages lead to tension during pharyngeal reconstruction, potentially increasing pharyngocutaneous fistula risk.Reference Do, Chung, Chang, Kim and Rho2,Reference Soylu, Kıroğlu, Aydoğan, Çetik, Kıroğlu and Akçalı23,Reference Aires, Dedivitis, Castro, Ribeiro, Cernea and Brandão24 However, studies by Cavalot et al., Erdag et al., and Saki et al., and our findings did not find a significant association between T stage and pharyngocutaneous fistula.Reference Cavalot, Gervasio, Nazionale, Albera, Bussi and Staffieri16,Reference Erdag, Arslanoglu, Onal, Songu and Tuylu20,Reference Saki, Nikakhlagh and Kazemi25 We propose that directly associating the T stage with tumour size may be misleading. Tumour size and its extension to nearby structures such as the base of the tongue or pyriform sinus might be more relevant risk factors than the T stage itself. Since our study excluded patients with such extensions, we cannot comment definitively on this parameter.
Several studies have linked prolonged surgical times and compromised local blood flow due to concurrent neck dissection to pharyngocutaneous fistula formation.Reference Wang, Xun, Wang, Lu, Yu and Guan5,Reference Virtaniemi, Kumpulainen, Hirvikoski, Johansson and Kosma26,Reference Dedivitis, Aires, Cernea and Brandão27 However, other studies have not found this association.Reference Boscolo-Rizzo, De Cillis, Marchiori, Carpenè and Da Mosto10,Reference Qureshi, Chaturvedi, Pai, Chaukar, Deshpande and Pathak22 Because all our patients underwent neck dissection, we could not analyse this specific relationship. Instead, we compared pharyngocutaneous fistula rates between patients who received radical neck dissection and those who received lateral neck dissection. Our findings identified radical neck dissection as an independent risk factor for pharyngocutaneous fistula formation, consistent with previous studies.Reference Lansaat, van der Noort, Bernard, Eerenstein, Plaat and Langeveld4,Reference Cavalot, Gervasio, Nazionale, Albera, Bussi and Staffieri16,Reference Galli, De Corso, Volante, Almadori and Paludetti18 However, some studies have not found this association.Reference Aydin, Taskin, Orhan, Altas, Ege and Yucebas3,Reference Boscolo-Rizzo, De Cillis, Marchiori, Carpenè and Da Mosto10,Reference Qureshi, Chaturvedi, Pai, Chaukar, Deshpande and Pathak22 Radical neck dissection is generally preferred for extensive lymph node metastasis. In these cases, removing metastatic lymph nodes takes considerably longer than standard elective neck dissection, and the amount of tissue resected is much greater compared to lateral neck dissection. Therefore, the explanation for this association between neck dissection and pharyngocutaneous fistula formation likely applies when comparing radical and lateral neck dissections as well.
Studies investigating the role of pharyngeal reconstruction methods in pharyngocutaneous fistula development have compared pharyngeal closure planes and suturing techniques. In the past, linear closure was considered superior to T-shaped closure because the intersection of the three lines was thought to be a high-risk area for pharyngocutaneous fistula.Reference Thawley28–Reference Shah, Ingle and Shah30 However, recent studies comparing T-shaped and vertical closures have not found significant differences in pharyngocutaneous fistula rates.Reference Chotipanich and Wongmanee9,Reference Qureshi, Chaturvedi, Pai, Chaukar, Deshpande and Pathak22,Reference Soylu, Kıroğlu, Aydoğan, Çetik, Kıroğlu and Akçalı23,Reference Saki, Nikakhlagh and Kazemi25 We think this might be attributed to the surgeon's experience rather than the specific pharyngeal closure plane itself.
Previous studies by Avcı et al. and Deniz et al. compared pharyngeal reconstruction techniques, finding significantly lower pharyngocutaneous fistula rates with continuous sutures (Zipper or Cushing) compared to simple interrupted sutures.Reference Deniz, Ciftci and Gultekin8,Reference Avci and Karabulut19 Similarly, Lemaire et al.'s research yielded consistent results.Reference Lemaire, Schultz, Vergez, Debry, Sarini and Vairel14 In our study, we compared T-shaped simple interrupted sutures with vertically modified Cushing sutures. The pharyngocutaneous fistula rate was significantly higher in the simple interrupted suture group; multivariate analysis confirmed pharyngeal reconstruction as an independent risk factor for pharyngocutaneous fistula.
These findings collectively suggest that continuous sutures are associated with lower pharyngocutaneous fistula rates, likely due to their ability to create a more watertight closure. However, some surgeons hesitate to use continuous sutures due to the concern that a single suture break could lead to a complete tear in the neopharynx. We acknowledge this concern and address it in our modified Cushing method by knotting the suture after every four stitches. This knotting technique minimises the risk of suture failure and subsequent tear, while still offering the benefits of a continuous suture. Based on these considerations, we think the modified Cushing method, with its continuous pattern and oblique transition, reduces the probability of pharyngocutaneous fistula compared to the simple interrupted suture technique.
• This study compared the effectiveness of a novel suturing technique, the modified Cushing suture to the traditional simple interrupted suture technique for pharyngeal reconstruction
• The findings demonstrated the superiority of the modified Cushing suture over the simple interrupted suture in minimizing pharyngocutaneous fistula development
• Since pharyngeal reconstruction techniques are current research topics for pharyngocutaneous fistula risk factors, this new technique contributes to the literature
This study has limitations inherent to its retrospective design. Information on parameters such as blood transfusion, antibiotic prophylaxis, and post-operative fever was unavailable from the hospital information-management system. Additionally, data on some factors such as comorbid diseases, pre-operative hypoalbuminemia, pre-operative anaemia, and smoking were limited due to missing information in a subset of patient records. Future studies should involve comparative studies with large numbers of patients. These studies could evaluate the effectiveness of the modified Cushing suture compared to other continuous sutures in preventing pharyngocutaneous fistula. Such research would provide valuable insights into the usability of modified Cushing among various continuous suture techniques.
Conclusion
Our findings demonstrate that pharyngeal reconstruction with simple interrupted sutures and radical neck dissection are independent risk factors for pharyngocutaneous fistula development. Based on the significantly lower pharyngocutaneous fistula rate observed, we suggest that the modified Cushing suture may be a preferable alternative to simple interrupted sutures for pharyngeal reconstruction.
Author contributions
Analysis and interpretation of data: YB, YA, GG, OU; drafting: YB, YA, GG; final approval of the version to be published: YB, YA, GG, OU.
Competing interests
The authors declare none.
Funding
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
