Factors Predicting Pharyngocutaneous Fistula in Patients after Salvage Laryngectomy - A Multicenter Collaborative Cohort Study

Presentation: AHNS21
Topic: Hypopharynx / Larynx
Type: Oral
Date: Wednesday, April 27, 2022
Session: 4:15 PM – 5:00 PM Oral Larynx
Authors: Conall W Fitzgerald, MD1; Joel C Davies, MD2; John R de Almeida, MD2; Sabrina Rashid, MPH2; Antoine Eskander, MD3; Eric Monteiro, MD3; Ximena Mimica, MD1; Marlena McGill, MPH1; Tim Mclean, MD1; Jennifer R Cracchiolo, MD1; Ian Ganly, MD1; Ahmed Teaima, MD4; Samantha Tam, MD4; Dongmin Wei, MD4; Ryan Goepfert, MD4; Jie Su, MSc5; Wei Xu, PhD5; Mark Zafereo, MD4; David P Goldstein, MD6; Marc A Cohen, MD1
Institution(s): 1Head & Neck Service, Memorial Sloan Kettering Cancer Center, New York, New York, USA; 2Department of Otolaryngology-Head & Neck Surgery, University of Toronto, Toronto, Ontario, Canada; 3Department of Otolaryngology-Head & Neck Surgery, Mt Sinai Hospital, University of Toronto, Toronto, Ontario, Canada; 4Department of Head & Neck Surgery, MD Anderson Cancer Center, Houston, Texas, USA; 5Department of Biostatistics, Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada; 6Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada


Introduction:

Pharyngocutaneous fistula (PCF) is a major morbidity associated with salvage total laryngectomy (TL). Understanding the factors that predict PCF is imperative to guide patient management. While numerous small retrospective cohort studies and a limited number of larger reports exist, no large, international collaborative studies on PCF in TL have been described to investigate risk factors for its development. We aim to assess the factors that predict PCF in patients after salvage TL in a multicenter collaborative retrospective cohort study.

Methods: Patients who underwent salvage TL for recurrent laryngeal cancer from January 1, 2000, to December 31, 2014 at 5 major academic centers in the US and Canada were included in this multicenter collaborative retrospective cohort study. Patient demographic data, dates of PCF, survival, response data, and detailed patient clinical data were collected. Rates of PCF were estimated by cumulative incidence function. Fine-Gray competing risk regression were conducted on time to PCF with death without PCF as competing risk. Univariable and multivariable analyses for factors associated with PCF were completed.

Results: In total, 550 patients who underwent TL were identified (mean [SD; range] age, 64 [10.4; 32-90] years; 465 [85%] men and 84 [15%] women). Primary treatment modality was radiation alone in 70% (n = 383) and concurrent chemoradiation in 30% (n = 167). Median radiation dosage was 66Gy (IQR 20-85Gy), with a median of 33fr (IQR 3-72). Free flap microvascular reconstruction (onlay or inlay) was undertaken in 25% (n = 139), while regional flaps were employed in 30% (n = 74). Primary tracheoesophageal puncture (TEP) was completed in 59% (n = 320).

Overall rate of PCF was 23% (n = 127). The median time to development of PCF was 2.9 weeks (interquartile range [IQR], 0.6-3.5 weeks). Surgical management of PCF was required in 43% (n=54) while the remaining 57% (n=73) required wound care alone. Rate of PCF varied by primary treatment modality [primary radiation, 20% (n=76); primary chemoradiation, 27% (n=40)]. Rate of PCF also varied with use of vascularized tissue in pharyngeal closure [free/regional flap onlay/inlay, 18% (n=25); no vascularized tissue/primary closure, 24% (n=98)]. Significant association between PCF following TL and advanced clinical local disease (T3 or T4), positive surgical margin, close surgical margin (<5mm), lymphovascular invasion and pre-treatment tracheostomy were identified on univariable analysis. There was no significant association on univariable analysis between PCF and treatment with chemoradiation (p=0.14) or use of vascularized tissue in reconstruction (p=0.12). Positive surgical margin (HR, 1.91; 95% CI, 1.11 to 3.29) was the only significant association on multivariable analysis.

Conclusion: This international multicenter retrospective cohort study highlights increased risk of PCF following salvage TL in patients with positive surgical margin. To our knowledge, this is the largest published dataset on PCF following TL, providing benchmark data across several high-volume academic head and neck oncology centers.