Elsevier

Journal of Voice

Volume 16, Issue 2, June 2002, Pages 289-302
Journal of Voice

Articles
Acoustical Aspects of Vocal Function Following Radiotherapy for Early T1a Laryngeal Cancer

https://doi.org/10.1016/S0892-1997(02)00100-5Get rights and content

Abstract

We evaluated acoustic voice characteristics of 18 male patients undergoing radiotherapy. The subjects were seen for voice assessment preradiotherapy and at 1 month, 3 months, 6 months, and 1 year following radiotherapy. A multidimensional voice analysis computer program (IVANS, Avaaz Innovations, 1998) was employed to evaluate measures of traditional frequency and amplitude perturbation as well as time-based and linear prediction (LP) modeledr “noise” parameters of the acoustic output in conjunction with perceptual judgments of overall vocal quality. The results indicate vocal deterioration of vocal function immediately following radiotherapy with gradual and significant improvement in acoustic and perceptual features over 9 to 12 months following the radiation treatment. Measures of glottal noise demonstrated higher sensitivity than frequency-based measures of voice perturbation, and with more consistent, less variable changes in acoustical voice output from the preradiation to the 12 month postradiation periods. Future research evaluating vowel type and acoustic perturbation measures with a larger sample of subjects over a longer time period seems warranted.

Introduction

Carcinomas of the head and neck region represent approximately 18% of all malignant tumors. Three percent of these are located within the laryngeal area,1 and 60%–65% of these tumors are squamous cell2 type that exist within the vocal folds.3 Laryngeal cancer primarily affects individuals who are middle age or older.4, 5 The disease is rare among nonsmokers and occurs most frequently among individuals who both smoke and have significant histories of alcohol use.4, 6

Most glottic carcinomas are surface lesions that may be visualized by direct laryngeal imaging. Early stage glottic carcinomas may invade through the mucosa of the vocal folds into the underlying superficial lamina propria which is observed as a tethering of the mucous membrane sliding over the underlying structures.4 The limited vibratory motion associated with small glottic tumors is best observed using videostroboscopy.7 Individuals suspected of having glottic carcinoma based upon laryngeal examination must undergo operative endoscopy and biopsy to confirm the diagnosis histologically.6 Following biopsy, the most commonly used clinical staging systems8 employ a classical three-component staging, i.e., “TNM” for describing tumors, node involvement, and metastasis.9

Therapeutic procedures for cancers of the vocal folds may involve surgery, radiotherapy, or a combination of both.9 Each method has strengths and weaknesses associated with its use.5, 10 Early diagnoses have led to the use of radiotherapy for many of the T1a (cancer of one vocal fold) patients seen in our clinics and is the method of choice within our center's care mandate.11 The schedule for radiotherapy treatment is usually “fractionated,” i.e., daily radiation treatments are given at a specific dose rate over a period of weeks to reoxygenate the tumor cells to increase their sensitivity to radiation destruction and to ensure repair and repopulation of “normal” cells which may be injured during radiotherapy.

The success of radiotherapy may be related to several factors. Favorable prognostic features include small lesion size, lack of subglottic extension, and absense of impairment of vocal fold mobility, a histologically well-differentiated tumor, and female gender. Less favorable signs include local recurrences relating to a geographical “miss,” c) consequences and complications associated with radiotherapy, including both acute conditions (i.e., erythema, oropharyngeal mucositis, mild laryngeal edema, and temporary hoarseness), and late effects, which relate to persistent tissue necrosis, impaired salivary function, edemas of the true and ventricular folds, and severe dysphonia.

Although not always present, the most common symptom of cancer of the vocal folds is persistent hoarseness.5 Hoarseness results from a vocal change that accompanies an alteration in the vibratory surface of the true folds due to the presence of the tumor. Otalgia and a dry cough may also be present as early symptoms.6

Assessment of vocal function preradiation and postradiation has been observed using both perceptual and objective measures. Typically, before radiotherapy the patient's voice has been subjectively rated as “hoarse, breathy, and/or strained.”12 Following radiotherapy, the voice has been described as “rough and hoarse” with a decrease in the breathiness and/or strain that was observed initially.12 Overall severity ratings of voice deviance decrease postradiation, but some patients still exhibit a significant amount of dysphonia when compared to control groups.13, 14

While subjective measurement of the patient's voice is an important component of the overall assessment, objective measures that allow evaluation of subtle changes in vocal function over time are preferred. Acoustical measures are of particular interest because they are noninvasive and easily obtained. A range of acoustical measures of vocal quality, including modal fundamental frequency,15 frequency range,16 frequency perturbation (jitter),16 modal intensity,15 intensity range,15 intensity perturbation (shimmer),16 and spectral voice characteristics (e.g., harmonics-to-noise ratio)16, 17, 18 are typically obtained before and following radiotherapy.19 These past investigations have indicated that acoustical voice assessments aid in describing the anatomical and physiological changes that occur as a result of medical, surgical, or radiological procedures.

Hirano et al20 evaluated the acoustic voice productions of 30 patients with voice disturbances (10 vocal fold cancers, 10 with unilateral polyps, and 10 with unilateral paralysis). Their investigation was designed to assess the efficacy of three specific perturbation parameters [i.e., pitch perturbation quotient (PPQ); amplitude perturbation quotient (APQ); and normalized noise energy (NNE)] in defining and differentiating the three groups of voice patients. In general, the investigators reported that while no one of the acoustic parameters was useful in differentiating the three groups, the three acoustic measures did prove beneficial in monitoring the effects of individual treatment for each of the three groups. In addition, all three acoustic parameters were positively correlated with perceptual judgments of voice quality.

Prosek et al21 assessed the relation between specific residue features of voice and the severity of voice symptoms as well as specific voice quality features. The investigators reported that the residue signals were useful in quantifying the amount or degree of severity of impairment. However, they noted that there was a lack of independence of voice quality features and features of specific acoustic residue measures.

Heeneman et al19 reported the use of a series of acoustical voice measures in patients with T1a cancers of the larynx before and after radiotherapy treatment. Their data indicated that there were differences in percent jitter and signal-to-noise ratio measures from before radiotherapy to 3–6 months following radiotherapy.

More recently, Dworkin et al22 reported phonation subsystem outcomes for 9 subjects over an average 15-month period following radiation treatment employing videostroboscopic, aerodynamic, acoustical, and perceptual measures. The investigators reported that many of the subjects demonstrated unique strategies in producing voice, and that no one physical measure was superior to another in detecting change over time.

Similarly, Verdonck-de Leeuw et al23 reviewed patients with early glottic cancer, managed radiologically, using multidimensional voice assessment consisting of videostroboscopy, phonetography, maximum phonation time, and phonation quotient. They reported that the patients' voice quality was poor before radiotherapy, improved following radiotherapy, and in 55% of these patients voice quality was comparable to control subjects in vocal function measures from 2 to 10 years postradiotherapy. Acoustical measures appeared beneficial in describing changes in voice production.

It appears that the usefulness of acoustic measures in detecting and tracking changes in the voice of individuals with vocal dysfunction may depend on the specific methods used to analyze the incoming voice signal. Techniques that demand accurate estimations of vocal fundamental frequency depend on the robustness of the signal, low noise levels, small modulations in vocal frequency and amplitude, and accurate methods of F0 extraction and tracking in order to operate efficiently. Such techniques are useful for mildly perturbed signals, but may not be robust enough to deal with very noisy and perturbed voice output.24 Recently, Parsa and Jamieson25 reported an investigation of performance of seven methods for F0 estimations with synthetic vowel waveforms. They concluded that the waveform matching (WM) algorithm was the most robust with pathological voice samples, remaining effective in the presence of perturbation and moderate levels of background noise. Such acoustical measures would appear useful in tracking change over time in patients treated for vocal fold cancer.

In view of the lack of data on the long-term effects of radiation therapy on acoustical voice characteristics in a subgroup of patients with T1a cancers of the vocal folds, and the need to carefully monitor vocal function relative to possible recurrence, the present investigation was designed to evaluate a variety of specific acoustical voice measures in adult male subjects receiving standard radiotherapy management.

Section snippets

Subjects

Subjects were 18 adult males with medically and histologically diagnosed carcinoma of one vocal fold (T1aNOMO). Subjects ranged from 37 to 85 years of age (mean 64 years). In all cases, the cancer was restricted to one vocal fold, was on the glottic plane, and had no anterior commissure involvement.

All subjects were followed by, and treated with, radiation therapy at the London Regional Cancer Centre (LRCC) as prescribed by the attending radiation oncologist. All subjects received 6,000 cGy of

Results

The acoustical data were subjected to several inferential, correlational, and descriptive analyses to evaluate the effects of radiation therapy over time and the variations in vowel type for each acoustical measurement over time. To that end, a multifactorial analysis of variance (MANOVA)(SPSS, Norusis)29 was completed, followed by univariate analyses of variance (ANOVA) for each measurement feature. In addition, correlational features and stepwise multiple linear regression analyses were

Discussion

This investigation was designed to study changes in the vocal function characteristics of 18 patients with cancer of one vocal fold (T1a) managed with external beam radiation over a one-year time period. Acoustical analyses and perceptual evaluations of the vocalization were undertaken at five measurement points before and subsequent to treatment.

The statistical trend noted in the present results, and depicted in the descriptive graphs, suggests that certain acoustical measures change

Conclusions

In summary, the present investigation was designed to evaluate the acoustical changes in vocal function over a one-year time period for 18 male patients treated with external beam radiation for Tla glottal cancer. The results of this study suggest that there are meaningful changes in a variety of acoustical vocal features over this time period, and that listeners' perceptions of these changes are consistent with a treatment model supporting the use of radiotherapy for simple glottal cancers.

Acknowledgements

We would like to recognize Karen Stenning for assistance in data analysis, Sarah Hawkins and Lori Holmes for aid in data collection, and K.G. Munhall for comments on the initial development of portions of the research. Financial support of portions of this research has come from CFI, ORTC, and NSERC (DGJ), CAA (VP), and Maude Catherine Taylor Cancer Research Fund (UWO) (HAL, KGM).

References (45)

  • PC Doyle

    Foundations of Voice and Speech: Rehabilitation Following Laryngeal Cancer

    (1994)
  • LW Davis et al.

    Controversy in the management of laryngeal tumors—radiation therapy perspective

  • G Duchesne et al.

    The nature of radiotherapy

  • ML Stoicheff et al.

    The irradiated larynx and voice: a perceptual study

    J Speech Hear Res.

    (1983)
  • AI Zablow et al.

    Radiotherapy T1 glottic carcinoma

    New Jersey Medicine

    (1989)
  • MD Kelly et al.

    Definitive radiotherapy in the management of stage I and II carcinomas of the glottis

    Ann Otol Rhinol Laryngol

    (1989)
  • T Murry et al.

    Changes in voice production during radiotherapy for laryngeal cancer

    J Speech Hear Disord

    (1974)
  • JJ Lehman et al.

    An objective assessment of voice production after radiation therapy for Stage 1 squamous cell carcinoma of the glottis

    Otol Head Neck Surg

    (1986)
  • RH Colton et al.

    Voice change after radiotherapy: some preliminary results

    Radiology

    (1978)
  • S Miller et al.

    Vocal changes in patients undergoing radiation therapy for glottic carcinoma

    Laryngoscope

    (1990)
  • H Heeneman et al.

    Vocal function following radiotherapy for early (T1) laryngeal cancer: a prospective study

  • N Hirano et al.

    Acoustical analysis of pathological voice: some results of clinical application

    Acta Otolaryngol

    (1988)
  • Cited by (16)

    • Voice Outcomes after Radiotherapy Treatment for Early Glottic Cancer: Long-Term Follow-Up

      2018, Journal of Voice
      Citation Excerpt :

      There is encouraging evidence for improved voice outcomes following RT. However, the research is largely limited to only 12 months8,13,14 and a 2-year follow-up.4,11 Due to favourable cure rates and subsequent long-term survivorship, more research is needed beyond 2 years to ascertain voice quality over a prolonged period of time following RT treatment.

    • Effects of voice rehabilitation after radiation therapy for laryngeal cancer: A randomized controlled study

      2014, International Journal of Radiation Oncology Biology Physics
      Citation Excerpt :

      After radiation therapy for laryngeal cancer, vocal function is impaired (1-3).

    • The harmonic and noise information of the glottal pulses in speech

      2014, Biomedical Signal Processing and Control
      Citation Excerpt :

      For HNR, the following three categories were used: 9–15 dB, 16–21 dB, and 22–27 dB. The first of these is typical for pathological voices while the second is characteristic to normal speech [33–35]. The last HNR range is related to voices which are highly periodic with a small amount of noise, such as the singing voice [36].

    • Voice outcome in patients treated for laryngeal cancer: Efficacy of voice rehabilitation

      2014, Journal of Voice
      Citation Excerpt :

      These results are in accordance with the present study. Worth noting is the deterioration of HNR values in the control group, which was not the case in the earlier mentioned studies.2,24,25 HNR has been shown to be a valuable variable in predicting voice quality with regard to hoarseness.27

    View all citing articles on Scopus

    This article is dedicated to the memory of Dr. H. A. Leeper, who passed away on May 24, 2001

    View full text