The Long-term Indwelling Tracheoesophageal Prosthesis for Alaryngeal Voice Rehabilitation

Objective  To analyze the initial experience at Oregon Health Sciences University, Portland, with the use of long-term indwelling tracheoesophageal voice prostheses.

Design  Retrospective case series.

Setting  Tertiary referral academic medical center.

Patients  Thirty patients undergoing speech rehabilitation after laryngectomy during a period of 18 months.

Intervention  Insertion of a long-term indwelling tracheoesophageal voice prosthesis.

Main Outcome Measures  Duration of use, complications.

Results  The mean duration of placement for a single prosthesis was 4.9 months (148 days), with a range of 14 to 330 days. Sixteen of the 30 patients encountered problems with leakage because of fungal colonization, the majority of which (15 of 16 cases) were solved with either oral or topical application of nystatin. Size matching in terms of prothesis length and tract length was critical, and problems of this nature were encountered in 11 of 30 patients. The incorporation of a second system of prostheses that offered an increased number of size options solved these problems in all of these patients. Ultimately, 27 of 30 patients were able to successfully wear these prostheses.

Conclusions  The indwelling tracheoesophageal voice prosthesis offers patients all the advantages of tracheoesophageal speech rehabilitation after laryngectomy without the inconvenience of frequent prosthesis changes. With careful attention to the details of fitting and care, it can be worn by the majority of patients successfully.

SINCE ITS introduction in the late 1970s, the technique of tracheoesophageal (TE) puncture and speech prosthesis placement has become the criterion standard for vocal rehabilitation after laryngectomy. Major innovation in placement and prosthesis design has occurred in both the United States and Europe, with a gradual overall improvement in the success rate of long-term patient use. The most recent innovation was the introduction of a long-term indwelling prosthesis intended to remain in position for months and relieve the patient of the inconvenience and problems associated with frequent prostheses changes. Few reports have been published, particularly in this country, on experience with the use of these prostheses.^1-3

The purpose of this study was to review the initial experience with the use of TE prostheses in a tertiary care, university-based head and neck oncology clinic at Oregon Health Sciences University, Portland.

Tracheoesophageal puncture as a method of vocal rehabilitation after laryngectomy has been used at Oregon Health Sciences University since the early 1980s. Consistent with the national trend in its use, we initially placed short-term patient-controlled prostheses as a secondary procedure. Beginning in the late 1980s, primary TE puncture assumed an increasingly prominent role in the treatment of patients undergoing laryngectomy; currently most patients have a 1.8-cm-long or 2.2-cm-long short-term patient-controlled prosthesis placed at the time of laryngectomy. Initial usage is begun approximately 2 weeks after the surgery when healing of the tracheostome and pharynx is ensured.

Issues related to the TE prosthesis in vocal rehabilitation are handled jointly by a speech and language pathologist and a head and neck surgeon, with the speech and language pathologist assuming the dominant role in management once the initial phase of healing is complete. Beginning in 1996, when the indwelling prosthesis first became available, the long-term indwelling TE prostheses were placed in a selected group of patients. These initial patients were those who, although functioning well with the shorter-term patient-controlled prosthesis, wished to be relieved of the inconvenience of frequent prosthesis changes or in whom issues related to difficulty in changing the prosthesis were a major factor in its overall care. As experience with the long-term prosthesis grew and the results were positive, there has been a shift toward placing the prosthesis earlier, either at the first postoperative visit after secondary TE puncture or once the TE tract has matured after primary TE puncture at the time of laryngectomy (usually 3 to 4 weeks). In both cases, the initial prosthesis placed intraoperatively is a short-term duckbill prosthesis of similar diameter (20F) because of greater ease of placement when compared with the configuration of the long-term prosthesis (Figure 1).

Although our initial experience with this prosthesis was exclusively with the Blom-Singer indwelling device (Inhealth Technologies, Carpinteria, Calif), in 1997 we began to use the Provox2 system (Atos Medical AB, Hörby, Sweden) selectively in an attempt to solve some of the problems encountered and because of the increased range of sizes that it offered.

Placement of the Blom-Singer indwelling prosthesis requires placement of a 22F dilator in the TE tract to allow insertion. The voice prosthesis is loaded into the gelatin cap loading tool, and a gelatin cap is placed on the posterior flange of the voice prosthesis to create a rounded edge for insertion (Figure 2, left). The gelatin-capped prosthesis is placed on the insertion stick. The dilator is removed and the prosthesis is quickly placed in the TE tract. The gelatin cap is allowed to dissolve (within 30 seconds), and the prosthesis is judged to be properly seated if the prosthesis can be easily rotated 360° within the tract and the posterior flange is seen nasoendoscopically to be open and flat against the esophageal wall. The tab can then be removed.

The Provox2 placement system is different (Figure 2, right). The Provox2 prosthesis is placed on the insertion stick and then placed into the loading tool. The loading tool is placed within the TE tract and the insertion stick is pushed through the loading tool to a specified mark on the stick. The loading tool is removed by pulling over the insertion stick while the insertion stick remains in place. Confirmation of placement is a flush faceplate against the posterior tracheal wall and anterior esophageal wall (by nasoendoscopy) and an easy 360° rotation within the TE tract. Once the prosthesis is confirmed to be in place, the tab can be removed. After placement, the patient is asked to voice to ensure that the vocal quality is at baseline and functional for communication. Instruction in cleaning techniques by means of the flushing pipette and the cleaning brush are given (Figure 3). Use of nystatin is also discussed.

The records of all patients in whom long-term indwelling TE prosthesis placement was attempted were reviewed. The type of prosthesis used, duration of placement, reasons for replacement or discontinuation of use, problems encountered, and solutions involved were noted.

Thirty patients underwent placement of a long-term indwelling prosthesis from September 1996 to January 1998 (Table 1). The majority of patients wore a Blom-Singer indwelling prosthesis (1.8-cm length in 13 and 1.4-cm length in 9). Twenty percent (6/30) wore a Provox2 device and 7% (2/30) reverted to wearing a patient-controlled device. The mean and median duration of placement for a single prosthesis were 4.9 months (148 days) and 4.4 months (135 days), respectively, with a range of 14 to 330 days.

Sixteen of the 30 patients encountered leakage of liquids through the voice prosthesis secondary to yeast colonization. This occurred with equal frequency in the Blom-Singer and Provox2 groups. A history of radiation therapy did not increase the frequency of this problem. The problem was solved in 15 of the 16 patients by treating them with or reminding them of the nystatin protocol (15m/L nystatin swish-and-swallow twice daily) (Table 2). In the last 6 months of analysis, the Provox2 cleaning brushes were introduced, which can be dipped in nystatin and run through the inner cannula of the prosthesis once a day as an alternative to taking the medication orally. One of the 16 patients, however, continued to have frequent recolonization and leakage. He ultimately resumed use of a short-term prosthesis of equivalent length and size to facilitate independent changing of the prosthesis.

Eleven of 30 patients initially had problems with sizing because of a TE tract length that fell between the 1.4-cm and 1.8-cm Blom-Singer indwelling prostheses sizes. A prosthesis that is too short causes granulation tissue to grow around the prosthesis, with subsequent aphonia, and a prosthesis that is too long causes leakage around the prosthesis. Five of the 11 patients had tracts that eventually settled to either a 1.4-cm or 1.8-cm length with expectant care alone. Six of the 11 were switched to a Provox2 8-mm (n=5) or 6-mm (n=1) prosthesis (Table 3). The intermediate length of these prostheses between the 1.4-cm-long and 1.8-cm-long Blom-Singer prostheses allows for flush seating of the prosthesis collars on both sides of the party wall, which was critical in extending the duration of placement by decreasing the amount of granulation tissue and leakage problems (Table 2).

One of 30 patients with sizing problems caused by granulation tissue (without obvious leakage) had such a frequent need for changing that he preferred to go to a shorter-term Blom-Singer device that could be changed independently. In 1 patient the tract was ultimately lost after 220 days with a 1.4-mm prosthesis. The reason for this remains unclear, as the patient lived out of state and was unavailable for follow-up at the time the tract was lost.

Although we did not objectively address voice quality with the long-term prosthesis as compared with the shorter-term device, patients subjectively reported a decreased "effort" with phonation, with a resultant increase in phonation duration. Vocal intensity or loudness was also subjectively improved over use of a 16F device.

Therefore, ultimately, 27 of the 30 patients were using the long-term prosthesis successfully and were pleased with the added convenience of not needing frequent changes. Two patients returned to the use of the patient-controlled short-term prosthesis because of the need for frequent changing with the long-term prosthesis despite all of the attempted solutions mentioned above and because they wanted to be independent of the need to return to the clinic for prosthesis changing. The tract was lost in 1 patient. Potential problems and their solutions are noted in Table 2.

The long-term indwelling TE prosthesis was designed to address the need for the patient to change the shorter-term prosthesis frequently. This need carries with it a certain amount of inconvenience and the potential for problems (tract loss, granulation tissue, and aspiration), especially for patients who have difficulties with manual dexterity, sight, or motivation. The long-term indwelling prosthesis allows patients who are "problem prone" to have their speech successfully rehabilitated with the TE puncture technique.

The long-term indwelling devices were not designed to improve speech over that seen with the short-term ones of similar diameter. The quality of speech is similar, and therefore patients should not be switched on this basis alone. Clearly, in the patient who is happy and independent in the use of a shorter-term prosthesis and having relatively few problems with it, the long-term prosthesis should be offered with caution. However, as our experience with the long-term prosthesis has increased, we are beginning to use it on a primary basis as we believe it offers advantages. Some patients can wear the short-term prosthesis for longer-than-usual periods. However, the average duration of use of the long-term prosthesis (4.9 months) in this study far exceeds the 6 to 8 weeks of use of the short-term prosthesis that we have seen in our population and that has been reported by others.^4,5

The financial aspects of the long-term device are also comparable. The clinic costs of the long-term devices are $80 for a Blom-Singer indwelling prosthesis and $182 for a Provox2 prosthesis. The 30-minute office visit with a speech pathologist required to change the device costs about $70. If 2 to 3 devices are required per year, the total cost will be $364 to $756, depending on the device used and frequency with which it is changed. The cost of a shorter-term patient-controlled device is $68, at approximately 6 devices per year ($408 per year, assuming no problems or office visits). If the short-term device must be changed by a professional because the patient is unable or unwilling to do so, the added costs of an office visit every 4 to 6 weeks can be substantial.

Our experience has taught us several things (Table 2). First, these prostheses are not completely maintenance free. The use of nystatin, either on a systemic basis or topically by daily application with a cleaning brush, is intermittently necessary in the majority of patients with either type of prosthesis to prevent fungal colonization and resultant leakage. The overall inconvenience of this is, however, believed to be less than that incurred by the need to change the shorter-term prosthesis more frequently, and the cost of nystatin (approximately $2.50 per day when used systemically) is minimal by comparison. Second, size does matter. The achievement of a length that allows a snug fit of both the tracheal and esophageal faceplates of the prosthesis without tension (with resultant granulation tissue) or play (with resultant leakage) is important. Our experience suggests that the incorporation of more than 1 system, in this case both the Blom-Singer indwelling and the Provox2 devices, is necessary to allow a sufficient variation in size to fit all tracts appropriately.

However, in experienced hands, these prostheses can be worn in the majority of patients (27 of 30) with a minimum of difficulty and inconvenience. The lack of need for frequent independent changing of the prosthesis increases the potential patient population to which TE puncture, the criterion standard for vocal rehabilitation after laryngectomy, can be offered and increases the desirability of this method of voice rehabilitation in general.

Accepted for publication September 10, 1998.

Presented at the Combined Meeting of the American Society of Head and Neck Surgery, Palm Beach, Fla, May 14, 1998.

Corresponding author: James Cohen, MD, PhD, Department of Otolaryngology/Head and Neck Surgery, Oregon Health Sciences University, PV-01, 3181 SW Sam Jackson Park Rd, Portland, OR 97201.