Voice-Related Quality of Life After Treatment of Laryngeal Cancer

Objective  To determine patient-perceived voice-related quality of life in patients treated with various methods based on the results of Voice-Related Quality of Life (VRQOL) and Voice Handicap Index-10 (VHI-10) questionnaires.

Design  The VRQOL and VHI-10 questionnaires.

Setting  University hospital.

Patients  One hundred thirty-seven patients who had received definitive treatment of laryngeal cancer were followed-up at Hokkaido University Hospital, Sapporo, Japan, and were alive with no evidence of malignancy at the time of the survey.

Main Outcome Measure  Patient-perceived voice-related quality of life based on the results of the VRQOL and VHI-10 questionnaires.

Results  The mean VRQOL scores for patients who had undergone radiotherapy (n = 63), chemoradiotherapy (n = 29), laser surgery (n = 14), or total laryngectomy (n = 27) as final treatment of laryngeal cancer were 92.6, 92.9, 85.5, and 68.4, respectively; the mean VHI-10 scores were 2.87, 2.34, 5.43, and 11.26, respectively.

Conclusion  The VRQOL and VHI-10 questionnaires are important in judging the overall effectiveness of treatment options for laryngeal cancer.

Radiotherapy and endoscopic laser surgery provide comparable cure rates for stage T1 glottic cancer.^1-3 In advanced laryngeal cancer, organ-preservation approaches that use combination chemoradiotherapy result in cure rates similar to those of primary laryngectomy with postoperative radiotherapy.^4-6 Hence, posttreatment morbidity is important in determining treatment selection. The assessment of morbidity by measuring function and quality of life (QOL) is also valuable for judging the overall effectiveness of newer treatment approaches and in helping to justify additional toxic effects. To our knowledge, few studies have reported on vocal function and QOL in patients after definitive treatment of laryngeal cancer using single voice-related QOL measures. The primary objective of this study was to examine patient-perceived voice-related QOL in patients treated with various methods on the basis of results of Voice-Related Quality of Life (VRQOL)⁷ and Voice Handicap Index-10 (VHI-10) questionnaires.⁸

Patients who received definitive treatment of squamous cell carcinoma of the larynx at all stages were recruited for the present study. After treatment, all patients were followed up at regular intervals at the Department of Otolaryngology, Hokkaido University Hospital, Sapporo, Japan. Surveys were completed at the follow-up visit between August 1, 2006, and May 31, 2007. Inclusion criteria for the study were freedom from disease and a minimum follow-up of 1 month since completion of final treatment.

One hundred thirty-seven patients (127 men and 10 women with a median age of 70 years at the time of the survey) were included in the study. Patient characteristics and treatment are given in Table 1. Patients had undergone either radiotherapy (≥60 Gy, and generally 65 Gy) with or without concurrent chemotherapy (carboplatin, 100 mg/m²/wk; or docetaxel, 10 mg/m²/wk; or cisplatin, 80 mg/m²/3wk), laser surgery under laryngomicroscopy, and partial or total laryngectomy as the final treatment of laryngeal cancer. Four patients who had undergone partial laryngectomy (either frontolateral or supracricoid) were excluded from further analysis because of the small sample in this subgroup.

The primary outcome variables were voice function as determined by validated QOL instruments: the VRQOL⁷ and the VHI-10.⁸ The VRQOL is a 10-item self-administered validated voice-outcomes measure. Scores are reported in 2 domains (social-emotional and physical functioning) and as a total score, all ranging from 0 to 100. A higher score indicates a better voice-related QOL. The VHI-10 scale, a short form of the Voice Handicap Index, is a 10-item self-administered validated instrument that measures patient disability as a result of voice disorders.^8,9 A higher index indicates poorer voice-related QOL. An adjusted VHI-10, ranging from 0 to 100, in which a higher index indicates better voice-related QOL, was also calculated to assess correlation with the VRQOL score. An auditory perceptual test using the GRBAS (grade, roughness, breathiness, asthenics, and strain) scale,¹⁰ which has been widely used since 1981,¹¹ was used when appropriate. We used the G-score (overall grade of voice) and scored on a 0 to 3 rating with 0 indicating normal; 1, slight; 2, moderate; and 3, severely dysphonic.

Pearson correlation analysis was used to calculate correlation coefficients between the VRQOL and adjusted VHI-10 or GRBAS G-scores. Comparisons between treatment groups and between tumor stages for outcomes of interest were assessed using the t test. All statistical analyses were performed using commercially available software (StatView version 5.0; SAS Institute Inc, Cary, North Carolina). A two-tailed P < .05 was considered statistically significant.

Completed VRQOL and VHI-10 questionnaires were received from 137 patients who received treatment of laryngeal cancer and who were free of disease at the time of the survey. Duration of follow-up after completion of treatment was 1 to 298 months (median, 38 months). In this cohort, 63 patients received radiotherapy, 29 received chemoradiotherapy, 27 underwent total laryngectomy, and 14 underwent laser surgery as the final treatment of laryngeal cancer. The relationship between time since completion of final treatment and total VRQOL and VHI-10 scores are shown in Figure 1 and Figure 2, respectively. Mean VRQOL scores (total, social-emotional domain, and physical functioning domain) and VHI-10 score for each treatment group are given in Table 2.

We examined the correlation between VRQOL and adjusted VHI-10 scores. As shown in Figure 3A, a highly significant correlation was observed between these 2 scores (r = .94). We also obtained G-scores from the acoustic perceptional GRBAS scale for a limited number of patients; these scores for each treatment group are given in Table 3. A comparison between G-score and total VRQOL score is shown in Figure 3B. A moderate inverse correlation was observed between these 2 scores (r = −.65).

To elucidate VRQOL outcomes after treatment of early glottic cancer, we compared VRQOL scores between patients with stage T1 glottic cancer treated with radiotherapy (n = 43) and those treated with laser surgery (n = 10) in a subanalysis of the patient cohort. In this subanalysis, patients who underwent laser surgery as first definitive treatment of glottic cancer were included in the laser surgery group. Despite significant differences in patient age and follow-up length, no significant difference in VRQOL scores was observed between these 2 treatment methods (Table 4).

In another subanalysis, we compared VRQOL scores among tumors classified as glottic T1a, T1b, or T2 and supraglottic T2 that were treated with either radiotherapy or chemoradiotherapy for stage 1 or stage 2 laryngeal cancer. Although a shorter time since treatment was noted in patients with T2 lesions, no significant differences in VRQOL scores were detected between patients grouped according to tumor classification (Table 5).

Inasmuch as many options for the treatment of laryngeal cancer provide comparable cure rates,^1-6 recommendations and patient counseling must consider the functional outcome after treatment. However, the best method to achieve an oncologic cure while minimizing adverse effects to the patient's voice remains controversial. Because, at least in part, there are problems associated with data collection, to our knowledge, few studies have examined voice outcomes using a single measure in all patients undergoing any of the various treatments of laryngeal cancer.

Auditory perceptual analysis of vocal outcomes after treatment of laryngeal cancer yields contradictory results. Although improved voice quality in patients undergoing radiotherapy has been reported,¹² results of another study found similar outcomes regardless of the type of treatment.¹³ Because the recording equipment, mouth-to-microphone distance, level of examiner training, and patient effort may influence analysis, methodologic issues limit the usefulness of perceptual voice measures.¹⁴

Acoustic voice parameters have also demonstrated inconsistent voice outcomes after treatment of laryngeal cancer.^15,16 Variability in the mouth-to-microphone distance, recording equipment, analysis software, and patient effort may explain the disparate results. Acoustic parameters also vary depending on whether vowels or speech samples are measured. In addition, acoustic variables reflect the sound of the voice but do not assess the patient's ability to communicate.^17,18 Most important, both perceptual and acoustic voice evaluations correlate poorly with patient-perceived vocal outcome.^15,19

Although the effect of treatment on the patient's voice remains the subject of debate, hoarseness and its functional consequences are a reality. How a patient regards voice impairment varies according to the individual. As opposed to imperfect quantitative analysis techniques, patient-based outcome assessments may be more pertinent to the patient when choosing treatment. Recent studies have used the VHI-10 or VRQOL questionnaire to assess treatment outcomes in laryngeal cancer.^20-24 Using these 2 voice-related validated questionnaires, we attempted to evaluate voice-related QOL in all patients treated with any of the various methods for laryngeal cancer.

In the present study, the mean (SD) VRQOL scores for patients who had undergone radiotherapy (n = 63), chemoradiotherapy (n = 29), total laryngectomy (n = 27), or laser surgery (n = 14) as the final treatment of laryngeal cancer were 92.6 (16.5), 92.9 (13.4), 68.4 (22.4), and 85.5 (14.9), respectively. Fung et al²⁴ reported that the VRQOL scores for healthy volunteers (n = 21), patients who had received chemoradiotherapy (n = 37), or patients who underwent total laryngectomy (n = 19) were 98.0 (3.9), 80.3 (20.8), and 65.4 (23.3), respectively. We observed a higher VRQOL score for patients who had received chemoradiotherapy than was reported by Fung et al.²⁴ This difference could have resulted because (1) the patients in the study by Fung et al²⁴ received a higher dose of cisplatin as chemotherapeutic agent, whereas most of our patients received a low dose of carboplatin or docetaxel and (2) the median time after final treatment was lengthier in our study.

In analysis of the association between time from the completion of the final treatment and VRQOL scores in patients who received either radiotherapy or chemoradiotherapy, we noted an increase in voice-related QOL in the first year after treatment. We also noted a gradual decrease thereafter, which was not observed in the laser surgery group. In addition, we found that a few patients who received either radiotherapy or chemoradiotherapy experienced a severe decrease in voice-related QOL because of late adverse effects of radiotherapy (Figure 1 and Figure 2).

Mean (SD) VHI-10 scores for patients who had undergone radiotherapy, chemoradiotherapy, total laryngectomy, and laser surgery were 2.87 (5.94), 2.34 (4.11), 11.26 (7.17), and 5.43 (5.75), respectively. Rosen et al⁸ reported that the VHI-10 scores for control subjects (n = 173), patients with Reinke edema (n = 27), and patients with recurrent nerve paralysis (n = 104) were 3.38 (5.65), 18.30 (10.97), and 25.72 (8.61), respectively. The VHI-10 scores for patients with laryngeal cancer compare favorably with those for patients with nonneoplastic lesions. There could be several explanations for these differences. Patients with laryngeal cancer may have different voice expectations than those with nonneoplastic disorders. Most of the patients with laryngeal cancer were smokers, and they may have had preexisting vocal fold changes that caused some degree of chronic dysphonia. Patients with benign lesions may be more likely to be heavy voice users and, consequently, more likely to be adversely affected by vocal fold abnormalities. These baseline differences in sensitivity to voice changes may influence voice-related QOL. After treatment of cancer, patient-perceived voice outcome may be influenced by the patients' satisfaction with being cured and having their voice preserved.²⁵ Patient voice expectations may be lower when they have neoplastic lesions compared with nonneoplastic voice disorders. In addition, voice quality may be less important than the ability to communicate for patients with laryngeal cancer. Regardless of the reason, the similar levels of voice-related QOL after radiotherapy and after laser surgery to treat stage Tl glottic cancer must be considered when counseling patients.

Many variables affect patient-perceived voice outcome after treatment of laryngeal cancer. Pretreatment factors included sex, tumor stage, and subsite of the primary tumor. Treatment factors included treatment group, that is, radiotherapy, chemoradiotherapy, total laryngectomy, or laser surgery resection. Posttreatment factors included time since completion of the final treatment. Other uncontrollable factors inherent to outcomes studies may be related to how patients view their voice after treatment. The sound of the voice, cultural influences, personality, marital status, job requirements, gastroesophageal reflux, tobacco use, posttreatment voice therapy, and age may all affect patients' perception of their voice, and none of these factors could be assessed.^19,26

In conclusion, we investigated voice-related QOL in patients who had received definitive treatment of laryngeal cancer based on the results of the VRQOL and VHI-10 questionnaires. Mean VRQOL and VHI-10 scores for total laryngectomy were 68.4 and 11.26, respectively, and were lowest among various methods, such as radiotherapy, chemoradiotherapy, laser surgery, and laryngectomy. Measures for voice-related QOL are increasingly important end points by which to judge overall effectiveness of standard and newer treatment methods.

Correspondence: Nobuhiko Oridate, MD, PhD, Department of Otolaryngology–Head and Neck Surgery, Hokkaido University Graduate School of Medicine, Kita 15, Nishi 7, Kita-ku, Sapporo 060-8638, Japan (noridate@med.hokudai.ac.jp).

Submitted for Publication: June 2, 2008; final revision received August 12, 2008; accepted September 1, 2008.

Author Contributions: Dr Oridate had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Oridate, Homma, Furuta, and Fukuda. Acquisition of data: Oridate, Homma, S. Suzuki, Nakamaru, F. Suzuki, Hatakeyama, Taki, Sakashita, Nishizawa, Furuta, and Fukuda. Analysis and interpretation of data: Oridate and Nishizawa. Drafting of the manuscript: Oridate. Critical revision of the manuscript for important intellectual content: Oridate, Homma, S. Suzuki, Nakamaru, F. Suzuki, Hatakeyama, Taki, Sakashita, Nishizawa, Furuta, and Fukuda. Statistical analysis: Oridate. Obtained funding: Oridate, Homma, Furuta, and Fukuda. Administrative, technical, and material support: Oridate. Study supervision: Homma, Furuta, and Fukuda.

Financial Disclosure: None reported.

Funding/Support: Drs Oridate, Homma, Nakamaru, Hatakeyama, Furuta, and Fukuda were supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, and Culture of Japan.

Previous Presentation: This study was presented as a poster at the Seventh International Conference on Head and Neck Cancer of the American Head and Neck Society; July 20, 2008; San Francisco, California.

This article was corrected online for typographical errors on 4/20/2009.