Tracheoesophageal voice prosthesis leakage can be intravalvular (more common) or periprosthetic (focus of this study). The authors studied the causes of periprosthetic leakage among 115 patients attending for voice prosthesis management (1374 clinic attendances) treated between December 2014 to December 2019 at the otolaryngology clinic of Fondazione Policlinico Universitario A Gemelli–IRCCS. Each patient was treated with the same algorithm until successful.
Therapeutic algorithm:
1. Inspection of local conditions (deep cleaning; prosthesis reallocation in situ).
2. Prosthesis replacement (shorter prosthesis; analogous prosthesis via overshooting; longer prosthesis).
3. Widening of the flanges (silicone ring behind tracheal flange; prosthesis with enlarged oesophageal flange).
4. Tissue augmentation (silicone/fat injection); purse string suture on fistula around prosthesis.
5. Fistula shrinkage (prosthesis removal and feeding tube insertion via fistula).
6. Fistula closure (definitive closure with surgical technique).
Over the study period, there were 330 leakages: 82/115 patients had leaks. Minor leakage (one-to-two leaks per patient, manageable with simple conservative procedures) 51/82 (62.19%) patients; major leakage (>2 events in a month +/- enlargement of the tracheoesophageal fistula) 24/82 (29.26%); and major fistula enlargement 7/82 (8.53%). Of the 82 patients, 25 patients (30.48%) had early leakage (<6 months from surgery), and 57 (69.51%) had late leakage (>6 months from surgery). Causes: too long prosthesis 35.15%; enlarged fistula 31.51%; granulation tissue formation or dislocated prosthesis 31.22%; and dilated, atrophic, infected, or necrotic fistula 2.12%. Risk factors: radiotherapy (preoperative or adjuvant) had no influence on attendances for leakage, however in those having leaks, there were higher rates of clinically relevant leakages or major fistula in patients who had primary irradiation. Primary vs. salvage total laryngectomy had no impact on major fistula formation or on attendances for leakage, but in those with leaks, significantly higher clinically relevant leakages in patients who had undergone salvage laryngectomy. No difference between results for primary vs. secondary tracheoesophageal puncture. Management: deep cleaning or prosthesis reallocation in situ was sufficient to resolve leakage 46 of 238 times (13.93% of all accesses). In all other cases, prosthesis was replaced. Insertion of a shorter prosthesis, performed 132 times, resolved leakage in 93 cases (28.18% of all accesses). Inserting an analogous prosthesis via overshooting replacement in seven cases solved 47 leakages (14.24% of all accesses). A longer prosthesis solved six accesses at the first attempt (1.81% of all accesses). A silicone ring behind the tracheal flange, performed 26 times, solved 18 cases (5.45% of all accesses). An enlarged flange, used 20 times, solved 14 cases (4.24%). Tissue augmentation: 14 silicone injections (six successful, 1.81% of all accesses) and eight fat injections (four successful, 1.21%). Purse string sutures always failed over the study period. In those cases, fistula shrinkage was pursued and succeeded in half the cases (2/4, 0.60% of all accesses). Two patients eventually had surgical closure.
