Vocal Rehabilitation after Total Laryngectomy: Provox Tracheoesophgeal Speech Valve Experience
George L. Charpied, M.S., C.C.C.-S.L.P.
Director, The Center for Voice & Swallow Disorders
Gregory A. Grillone, M.D., F.A.C.S., Vice-Chairman
Robert W. Dolan, M.D., F.A.C.S.,
Director, Lahey Clinic, Burlington, MA
Abstract:
Surgical efforts to provide voice production after total laryngectomy has had varying degrees of success. Tracheoesophageal speech with valve prosthesis has become an important method of voice rehabilitation with refinements such as the Blom-Singer Duckbill tracheoesophageal prosthesis and the Panje button prosthesis worldwide.
We report here, data collected from three years experience in our department on complication rates, device lifetime, replacement maintenance procedures, cost, and the efficacy of the Provox prosthesis when compared with the Blom-Singer voice prosthesis.
Subjects included 11 males (mean age = 60) and three females (mean age = 54). Primary and secondary valve placement procedures were considered. The time interval for secondary procedure placement of tracheoesophageal speech valve ranged from four to 24 months (mean = 14 months). In all except one case, patients underwent total laryngectomy with unilateral or bilateral neck dissection and radiation therapy for treatment of laryngeal carcinoma.
The single most common complication for the Blom-Singer was extrusion or accidental removal (85%), followed by improper length or French size (62%), valve leakage (38%) due to clogging, improper valve orientation, or poor maintenance. For the Provox, extrusion and improper sizing fell to zero percent and valve leakage was 15%. Speech and acoustic data revealed functional voices in all but one patient.
The Provox produced better results for durability, ease of maintenance, reduced strain when speaking and voicing characteristics. In our experience prosthetic shunt tracheoesophageal speech is the preferred method of voice rehabilitation. Based on its 'near' normal intelligibility and fluency and shallow learning curve, the Provox was clearly the superior device.
Introduction:
With the introduction of laryngectomy for treatment of laryngeal cancer in the late 1800's by Watson, 1866, and Billroth, 1874, the loss of ability to communicate verbally has been the single most disabling consequence of this procedure (Grussenbauer, 1874; Weissenbuch & Albers, 1993).
Since Billroth, surgical intervention to initiate voice production after total laryngectomy has had varying degrees of success. Rehabilitation efforts have focused on establishing voicing with the artificial larynx or esophageal speech (Calcaterra, 1985).
Alaryngeal speech through an artificial larynx, such as the Servox electrolarynx, has the disadvantage of mechanical sound production and device handling problems. Additionally, the absence of pitch or loudness control and the visually distracting nature of the mechanical electrolarynx with alaryngeal speech has proven effective, but less than desirable for interactive communication.
Esophageal speech is considered the “method of choice” for rehabilitation after total laryngectomy. Some authors estimate that 30-70% of laryngectomies use esophageal speech as their primary means of communication (Panje & van Demark, 1985). However, only 25% of those patients achieve functional esophageal speech proficiency (Hilgers & Balm, 1993). Esophageal speech has the advantage of not requiring digital closure of the stoma but the disadvantage of a long and difficult, complicated learning process. Compared with electrolarynx speakers, esophageal speech has the inherent limitations of less words per minute, reduced phrase length, reduced pitch and loudness, 'stoma blast' and 'clunking' on air intake (Qi & Weinberg, 1995; Robbins, 1984).
Tracheoesophageal speech using valve prosthesis has become an important method of surgical voice rehabilitation. Refinements of the Blom-Singer Duckbill tracheoesophageal prosthesis and the Panje button prosthesis have been reported worldwide (Panje, 1979; Blom & Singer, 1981; Hilgers, et al., 1993).
Zwitman and Calcaterra believed a prosthetic device could take advantage of esophageal distensability in which voluntary control of the cricopharyngeus muscle of the upper esophagus might be employed (Moon, 1987; Robbins, et al., 1988; Hirose, 1996; Zwitman and Calcaterra, 1973). Recently, several prosthetics have been developed for tracheoesophageal puncture, performed as a primary procedure during laryngectomy, or as a secondary procedure at a later stage after cancer treatment.
Our study compared two types of in-dwelling speech valves, the ProVox and Blom-Singer devices. In on our experience, prosthetic shunt tracheoesophageal speech is in accord with other findings as the preferred method of voice rehabilitation. We found a 'near' normal intelligibility, fluency and a shallow learning curve than with the electrolarynx or esophageal speech (Miralies & Cervera, 1995). Problems encountered during regular use of tracheoesophageal prostheses are related to high resistance to airflow through the prosthesis and esophageal segment, spontaneous extrusion during coughing, sneezing, or cleaning and limited lifetime.
The development of sturdier low resistance valve prostheses with particular attention to the valve part of the prosthesis was an issue to be addressed by others (Pauloski, et al., 1989; Hilgers, et al., 1993; Hilgers & Balm, 1993) as it affects ease of use. Although unilateral myotomy and/or selective unilateral neurectomy of the pharyngeal plexus have reduced mean intra-tracheal voicing pressure and muscle tone through the pharyngoesophageal segment, development of an improved prosthesis that reduces the need for additional surgery is more desirable (Singer & Blom, 1981).
Another consideration is deterioration of the silicon prosthesis by fungal colonization and erosion, which causes valve dysfunction and leakage of esophageal contents into the tracheal airway. These concerns were not previously addressed with available technology (Nataragan, et al., 1994).
This report presents our data from three years experience and includes comparative data on complication rates, device lifetime, replacement and maintenance procedures, and cost in our department. We were interested in the efficacy of the Provox prosthesis and began using the Provox self-retaining, low-resistance prosthesis for vocal rehabilitation after total laryngectomy beginning in June, 1995.
Patients and Methods:
Between June of 1995 and June of 2002, 13 patients underwent primary and secondary tracheoesophageal puncture with insertion or a Provox and the Blom-Singer voice prosthesis. This study included 13 patients; 10 males (mean age = 60) and 3 females (mean age = 54).
Primary puncture was performed in three patients (all males) and secondary procedures or valve replacement were performed in eleven patients (three females and eight males). The time interval for secondary procedure placement of tracheoesophageal speech valve, regardless of type, ranged from 4 to 24 months (mean = 14 months). With the exception of one patient, treatment for laryngeal carcinoma consisted of total laryngectomy with no neck dissection, or a unilateral or bilateral neck dissection as the primary surgical treatment of laryngeal carcinoma, with adjuvant radiation therapy. One patient underwent a partial laryngectomy with construction of a neo-vocal fold, which failed, necessitating a salvage total laryngectomy. Thirteen patients received post-operative radiotherapy (50-70 Gy). In total, 32 Blom-Singer and 8 Provox prosthesis were inserted in our subjects.
Primary or secondary unilateral cricopharyngeal and constrictor myotomy to prevent problems with hypertonic pharyngoesophageal segment were not performed. Primary insertion of the Provox speech valve was done at the time of laryngectomy as defined by Hilgers, et al., 1993. Secondary tracheoesophageal puncture and insertion of the Provox speech valve was performed endoscopically with a Haslinger handle after insufflation test revealed “within functional limits” voicing capability, as described by Blom and Singer. The Provox offers an anterior approach insertion obviating the need for the more complicated “pull through” technique. In one patient, primary insertion of the speech valve was complicated by severe swallowing impairments post-surgically and the Provox was removed. In all cases, speech valve replacement was performed as an outpatient procedure.
Contraindications for tracheoesophageal puncture were varied and included patients not wanting additional surgical procedures or non-motivated patients. Also patient with subglottic tumor extension below the fourth tracheal ring, those in poor general health, with persistent swallowing problems, impaired pulmonary function, poor mental or digital dexterity, and those with poor visual acuity or mental disability.
All patients received extensive pre- and post-operative in-hospital or in-clinic counseling regarding all available voice and rehabilitation options. Manuals were provided to patients and their families explaining all procedures, best estimates of time frames for rehabilitation, possible complications, and a video-tape demonstrating alaryngeal speech types.
Voice rehabilitation was performed by a speech language pathologist in conjunction with a surgical team. Patients were trained to acquire esophageal speech or use the electrolarynx in addition to the tracheoesohageal speech valve in order to provide a backup mode of communication. A protocol was developed for evaluation of alaryngeal speech based on the 'Third International Congress on Voice Prosthesis' held in Groningen, Germany, 1986 (Schutte, et al., 1986). The modified protocol was divided into three sections:
- patient demographics, including cancer type and site and surgery type,
- natural history of the tracheoesophageal speech valve, including type, time frames, and complications, and
- acoustic and speech characteristics, including intelligibility, voicing skill, acoustic data and patient satisfaction.
Results:
Complications
Follow-up period ranged from 4 to 24 months (mean = 14 months). One patient developed a recurrence of his cancer and died. Another patient’s speech valve was removed due to swallowing difficulties. The single most common complication for both valve types was extrusion (85%), improper length or French size (62%), valve leakage (38%), clogging due to excessive secretions, improper valve orientation or poor maintenance (31%). In some patients stoma or upper esophageal stricture requiring stomaplasty or dilatation and granuloma formation over the esophageal side of the speech valve (15%) were also encountered. Finally, there were reports of increased gastroesophageal reflux (23%) paroxysms of coughing (15%) in some patients. See Table 1.
Table 1. Summary of Speech and Acoustic Data.
Key: Valve - valve type where P = Provox and B-S = Blom-Singer; Syl/Br = estimate of syllable per breath from tape recorded speech sample; MPT = maximum phonation time; WPM = estimate of word per minute rate from tape re-corded speech sample; seal = judgment of effective consistency of digital stoma seal for speech; ODC = judgment of on-demand speech consistency and reflects patient ability to speak when he or she desires; Strain = combination speech pathologist's judgment and patient's report of the amount of effort employed to obtain speech; Intel. = judg-ment of patient's speech intelligibility on a scale of 1 (poor) to 5 (excellent) during three successive therapy ses-sions; F0 = patient's fundamental frequency, as derived from a digitally recorded sustained /a/, then analyzed using Computer Speech Laboratory 5.0 [Kay Elemetrics, Pine Wood, NJ]; DOV = degree of unvoicedness in percent de-rived from CSL; DynRng = estimate of loudness dynamic range in decibels [dB] from a recorded speech sample analyzed using CSL; HNR = estimate of harmonic - to - noise ratio from recorded connected speech sample.
Rating Speech
The simple rating system of voice quality and acoustic character used in this study was revealing. Our patients had a mean score of 60% for proper on-demand stoma seal and a mean score of 4.0 (S.D. = 0.816) for intelligibility. Timing features such as; syllables per breath, phrase length, maximum phonation time and words per minute (rate) were consistent with other reports. Our patients mean for syllables per breath was 15.7 (S.D. -5.37), maximum phonation time was 5.4 seconds (S.D. = 3.034) and words-per-minute (rate) was 118.285 (S.D. = 47.633), all of which were determined to be within the range of data previously reported (Debruyne, et al., 1994). Acoustic features such as; voicing, fundamental frequency, loudness dynamic range, degree of unvoicedness and harmonic-to-noise ratio were tested. Our patients mean for fundamental frequency was 188.500 Hz (S.D. = 93.879), degree of unvoicedness of 61.428% (24.130), loudness dynamic range of 37.4 dB (S.D. = 17dB), and harmonic-to-noise ratio of -7.404 (S.D. -- -3.002) differs only slightly from previously reported data (Zwitman & Cal-caterra, 1974; Debruyne, et al., 1994; Qi & Weinberg, 1995). See Table 1.
Speech and acoustic data
Speech and acoustic data revealed functional voices in all but one patient. Syllables per breath ranged between 7 and 22 (mean = 15.7), maximum phonation times ranged between 2 and 10 seconds (mean = 5.4), words per minute (rate) ranged between 40 and 173 (mean = 118.3), fundamental frequency (F0) ranged between 115 and 397 Hz (mean = 199.6), dynamic range was between 61 dB and 21 dB (mean = 37.4) and harmonic to noise ratio was between +3.8 dB and -50 dB (mean = 7.4). Our patient population had score ranges of 1 to 5 for on-demand stoma seal with a mean of 4.0 (S.D. = 0.8), intelligibility scores of 2 to 5 with a mean of 4.0 (S.D. = 0.7). Timing features such as syllables per breath, phrase length, maximum phonation time and word per minute rate were consistent between valves types, and were consistent with other reports. Our patient’s means for syllables per breath of 15.7 (S.D. = 5.4), maximum phonation time of 5.4 seconds (S.D. = 3.0), words-per-minute (rate) of 118.3 (S.D. = 47.6) were within the range of data already reported.
Discussion:
Vocal rehabilitation after total laryngectomy has been a major concern to head and neck surgeons since the first total laryngectomies of Watson and Billroth. Grussenbauer described the technical aspects of the application of Cerzny's vocal cannula, which Grussenbauer called an artificial larynx, in the attempt to rehabilitate speech after total laryngectomy (Grussenbauer, 1874). Since the 1870s, aspiration and stenosis, and to a lesser degree tissue breakdown in the form of diverticulum or necrosis, have been seemingly intractable complications. After nearly 100 years of experimentation and failure, voice rehabilitation after total laryngectomy was achieved with esophageal speech and/or the artificial electrolarynx as sound sources (Zwitman & Calcaterra, 1973).
Beginning in the early 1980's truly successful tracheoesophageal puncture and insertion of a plastic prosthetic device such as the Blom-Singer and Panje tracheoesophageal speech prosthesis with their pulmonary driven mechanism established a new age of speech rehabilitation (Izdebski, et al., 1994). Since that time, many more prosthetic devices have been developed.
In our experience the indwelling ProVox type prosthesis has advantages over other speech valves. The advantages include the benefit of patients leaving the hospital two to three weeks after total laryngectomy with a useful and intelligible voice, that likely translates into better psychosocial adjustment (Baugh, et al., 1990) due to primary prosthetic voice placement. The influence of early prosthetic voice rehabilitation on psychosocial rehabilitation has yet to be thoroughly investigated and further study is necessary. Also there are reports that secretion production and olfaction are improved with speech valves prosthesis (Hilgers, et al., 1993; Lichtman, et al., 1995).
The voice prosthesis does not prevent the patient from learning alternative forms of laryngeal speech. In fact, the patient is expected to learn either the use of an electrolarynx or esophageal speech as an adjunct communicative method in the event the prosthetic speech valve becomes incompetent. Lastly, cleaning was easy and effective valve longevity was improved.
Approximately three-quarters of our patients acquired useful mechanical electrolarynx voicing, while one-quarter acquired esophageal speech in addition to the ProVox valve. These variations from the literature appear to have two basis. First, our clinical practice aims at providing patients with the best possible voice rehabilitation and the indwelling prosthesis has proven itself to be the prosthetic of choice. Second, although some estimate up to sixty percent of patients acquire esophageal speech, a better estimate would take into account the usefulness of the esophageal speech. Esophageal speech usefulness has been shown to be about 25% (Robbins, 1984).
Radiotherapy for primary voice prostheses insertion patients does not appear to influence longevity or success of voice rehabilitation (deCarpentier, et al., 1996) except for the inevitable deterioration of voicing during the final weeks of post-operative radiation. No other increases in prosthesis-related complications were documented.
Unintentional removal or extrusion of the valve from the tracheoesophagel puncture is the most common complication in our experience. With previous Blom-Singer type speech valves it occurred during coughing bouts or during manual manipulation, such as cleaning. For us, accidental removal was the single best reason to look for an alternative. However, the Provox presented us with the problem of extrusion into the esophagus. This was quickly resolved as our experience improved and we learned to properly size patients when fitting them with the device. With the Blom-Singer, extrusion remained an ongoing problem.
The second most common complication is valve leakage (Van Den Hooger, et al., 1996). However, the valve is not the reason for leakage. Colonization of the valve leaflet and diameter of the valve chamber just anterior to the leaflet by Candida albicans microfloral fungus has been documented (Natarayan, et al., 1994). These tenacious fungal colonies cause the valve leaflet to remain open as colonies accumulate on it s perimeter. Having patients clean with a brush moistened in a dilute solution of hydrogen peroxide four times daily resolved the problem satisfactorily for both speech valves. We also examined the efficacy of anti-mycotic prophylactics such as Nystatin Swish-and-Swallow and determined that a reduction in the rate of voice prosthesis leakage was accomplished.
Radiotherapy for primary voice prostheses insertion patients does not appear to influence longevity or success of voice rehabilitation (deCarpentier, et al., 1996). However, there is an inevitable deterioration of voicing during the final weeks of post-operative radiation, with no increase in prosthesis-related complications.
References:
Hilgers, F.J.M. and A.J.M. Balm (1993) Long-term results of vocal rehabilitation after total laryngectomy with low-resistance, indwelling Provox voice prosthesis system. Clinical Otolaryngology 18:517-523.
Lichtman, S.W., I.L. Birnbaum, M.R. Sanfilippo, J.T. Pellicone, W.J. Damon and M.L. King (1995) Effect of tracheostomy speaking valve on secretions, arterial oxygenation, and olfaction: A quantitative evaluation. Journal of Speech and Hearing Research 38:549-555.
Miralies, J.L. and T. Cervera (1995) Voice intelligibility in patients who have undergone laryngectomies. Journal of Speech and Hearing Research 38:564-571.
Panje, WR and DR VanDemark (1985) Experience with the Panje voice prosthesis. In: Surgery_ of the Larynx, pp. 385-396, Ed.s Bailey, BJ and EF Biller, W.B. Saunders, Philadelphia.
Pauloski, B.R., H.B. Fisher, G.B. Kempster and E.D. Blom (1989) Statistical differentiation of tracheoesophageal speech produced under four prosthetic/occlusion speaking conditions. Journal of Speech and Hearing Research 32:591-599.
Qi, Y. and B. Weinberg (1995) Characteristics of voicing source waveforms produced by esophageal and tracheosesophageal speakers. Journal of Speech and Heating Research 38: 536-548.
Robbins, J, J, Christiansen and G. Kempster (1986) Characteristics of speech production after tracheoesophageal puncture: Voice onset time and vowel duration. Journal of Speech and Heating Research 29:499-504.
Robbins, J (1984) Acoustic differentiation of laryngeal, esophageal and tracheoesophageal speech. Journal of Speech and Hearing Research 27: 577-585.
Schutte, H.K, E.F.M. Bors, G.H.A. van Doer, G.L.J. Neiboer, and A.A. Annyas (1986) Evaluation of speech with and without a Groningen type voice button. In: Speech Restoration Via Voice Prosthesis. pp. 135-138. Ed. I.F. Herrmann, Springer-Verlag, Berlin.
Singer, M.I. and E.D. Blom (1981) Selective myotomy for voice restoration after total laryngectomy. Archives of Otolaryngology - Head and Neck Surgery 107:670-673.
Singer, M.I. (1996)Total laryngectomy with voice preservation. In: The Larynx: A Multidisciplinary_ Approach pp. 541-547, Ed. Fried, MP, CV Mosby, Boston.
Singer, M.I. and E.D. Blom (1995) Vocal rehabilitation with prosthetic devices. In: Diagnosis and Treatment of Voice Disorders, pp. 367-384, Ed.s Rubin, J.S., R.T Sataloft, G.S. Korovin and W.J. Gould, Igaku-Shoin, New York.
Van Weissenbruch, R and FWJ Albers (1993) Vocal rehabilitation after total laryngectomy using the Provox voice prosthesis. Clinical Otolaryngology 18:359-364.
Wong, S.H.W., C.H. Cheung, A.P. Wing, W.K.Ho and W.I. Wei (1997) Assessment of tracheoesophageal speech in a tonal language. Archives of Otolaryngology - Head and Neck Surgery 123:88-92.