Key points

Introduction

Transoral laser microsurgery (TLM) has become the mainstay in the surgical treatment of early-stage (TisT1T2) glottic cancer with excellent oncologic outcomes, comparable to nonsurgical options ( Table 1 ). When evaluating different treatment options for early glottic cancer, factors such as patient age and comorbidities, treatment availability, cost and duration, and risk of second primary cancer should be taken into consideration. However, because of the favorable prognosis of these tumors, long-term voice quality and quality of life for these patients is the major factor in selecting the type of treatment.

The approach to evaluating the dysphonic patient is described in another article by S. Dailey and colleagues in this issue. This article covers the current literature regarding subjective voice quality and voice-related quality of life and studies with objective acoustic and aerodynamic results after treatment of early glottic cancer with TLM.

Introduction

Transoral laser microsurgery (TLM) has become the mainstay in the surgical treatment of early-stage (TisT1T2) glottic cancer with excellent oncologic outcomes, comparable to nonsurgical options ( Table 1 ). When evaluating different treatment options for early glottic cancer, factors such as patient age and comorbidities, treatment availability, cost and duration, and risk of second primary cancer should be taken into consideration. However, because of the favorable prognosis of these tumors, long-term voice quality and quality of life for these patients is the major factor in selecting the type of treatment.

The approach to evaluating the dysphonic patient is described in another article by S. Dailey and colleagues in this issue. This article covers the current literature regarding subjective voice quality and voice-related quality of life and studies with objective acoustic and aerodynamic results after treatment of early glottic cancer with TLM.

When to evaluate voice after transoral laser microsurgery

The thermal effects of TLM with postoperative inflammation and localized edema generally subside within 1 to 2 months and lead to the formation of scar tissue that can evolve over time. This inflammation and scarring can lead to variable voice quality during the first few months postoperatively. Definitive voice evaluation should be undertaken only when the voice quality has stabilized. Several studies attempted to ascertain the appropriate time after which the voice is stable and thus definitive voice quality is evaluable.

In a prospective cohort study, voices of 106 male patients with T1 tumors treated with type II subligamental cordectomy were recorded preoperatively, then at 3, 6, 12 and 24 months postoperatively. The objective acoustic measurements of jitter, shimmer, and normalized noise energy returned to within the normal range after just 3 months, but the voice fundamental frequency remained in a higher range than normal (140 Hz) even after 24 months.

In a smaller and more heterogeneous study, prospectively evaluating patients preoperatively and at 1, 3, 6, and 12 months, Chu and colleagues found that the subjective evaluation (grade, roughness, breathiness, asthenia, and strain [GRBAS scale]) and the objective acoustic and aerodynamic parameters improved and then stabilized at 6 months for limited and more extensive resections. In a prospective cohort study, Lester and colleagues found that for the 19 T1 tumors treated with TLM, acoustic and aerodynamic measurements were abnormal 3 months postoperatively but returned to preoperative values after 12 months. Other studies have shown voice stabilization after 12 months.

Finally, Spielmann and colleagues, published a systematic review in 2010 encompassing 15 studies comparing voice outcomes in patients treated with TLM versus RT. The authors cite 1 study evaluating voice after 3 months, 3 studies with evaluations after a minimum of 6 months, and the 11 remaining studies evaluating voice after 12 to 24 months or more.

It appears, then, from the literature, that definitive voice evaluation should not be performed before 6 months postoperatively and that it may be more appropriate to evaluate voice after an interval of 12 months because of some heterogeneity in the healing process, possibly affected by the extent of TLM.

Interestingly, in a self-oscillating physical model of the vocal folds, Mendelsohn and colleagues found that adding a scar to the model of the vibrating vocal fold, voice was improved compared with the model of vocal fold resection without scarring. Thus, in real life, it is possible that a certain rigidity of the vocal fold, obtained only after a period of healing, will affect the definitive vocal outcome, either by an acoustic or an aerodynamic mechanism.

How is voice subjectively after transoral laser microsurgery?

Subjective voice results have been studied in numerous publications of prospective and retrospective cohorts. The methods of subjective evaluation vary among publications, with patient self-reporting or blinded or nonblinded expert reporting. The scales vary from visual analog scales to the GRBAS scale, the most largely used. Many of these publications have been retained in 4 meta-analyses and 1 systematic review. Only one recent prospective, randomized study comparing voice results after TLM with voice after RT has been published.

In a cross-sectional study of 42 consecutive male patients, evaluated 6 to 48 months postoperatively, Vilaseca and colleagues found that definitive subjective voice quality was correlated to the amount of tissue resected, that is, to the type of cordectomy performed. After a type I cordectomy, two-thirds of patients had a “normal to near-normal” voice, as rated by the otolaryngologist, and, after types II and III, 55% of patients had a normal or near-normal voice, whereas only 25% had a normal or near-normal voice after a type V cordectomy. Thirty-one percent of these latter patients were judged to have severe dysphonia. Compared with normal controls, all of the GRBAS categories, as judged by a speech therapist, were significantly worse after TLM, and particularly for the grade, breathiness, and asthenia categories. On a scale from 0 to 3, type I, II, and III cordectomies resulted in GRBAS scores of approximately 1, whereas more extended cordectomies resulted in GRBAS scores of greater than 2.

Physiologically, a near-normal voice after a type I cordectomy may be explained by the preservation of the superficial lamina propria (Reinke’s space), which preserves the mucosal wave, once the mucosa has healed. Deeper cordectomies, however, remove the superficial lamina propria at the tumor site. When the mucosa regenerates, the scar tissue is fixed to the underlying tissues with little or no vibratory qualities.

Similarly, Aaltonen and colleagues, in a prospective study of 32 patients 2 years after TLM, found a median GRBAS score of 1.5. Nineteen percent of their patients had a G score of 0, whereas 7% had a G score of 3. Resection of the anterior portion of the vocal fold was correlated with worse vocal outcomes in this study. Several other studies found the same correlation between the extent of the cordectomy (types V and VI) and a worse subjective vocal outcome. For exclusively T1a lesions treated with TLM, Czecior and colleagues found that 54% of patients had a G score of 0 to 1, 78% an R score of 0 to 1, 96% a B score of 0 to 1, 99% an A score of 0 to 1, and 58% an S score of 0 to 1. Compared with preoperative voice, Aaltonen and colleagues found a significant improvement in average self-reported hoarseness on a visual analog scale of 0 to 100 after 6 months (50/100 vs 59/100 preoperatively), improving further after 12 months (43/100). Thus, most patients will have normal to near-normal voices after TLM, particularly after type I, II, and III cordectomies. Poorer subjective voice quality is to be expected with more extended cordectomies.

How is voice-related quality of life after transoral laser microsurgery?

Some voice-related quality of life studies include studies based on responses to general questions such as “effect of voice on daily living” measured with visual analog scales or Likert scales, but the most widely used metric for evaluating the daily handicap associated with voice quality is the Voice Handicap Index (VHI), either in its initial 30-question version, or its reduced 10-question version. The questionnaire evaluates physical, functional, and emotional domains of voice handicap. Reported scores on the VHI-30, where 120 represents the worst most incapacitating voice, range from 11.5 to 29.2 for T1a glottic cancers treated with TLM. For comparison, the average VHI in the case of benign vocal fold lesions is approximately 26; for spasmodic dysphonia treated with botulinum toxin it is 22 and after medialization thyroplasty the score is 28. As for subjective voice analysis, the degree of voice handicap is correlated to the depth and extent of the cordectomy. Peretti and colleagues found an average VHI of 6 for superficial type I and II resections, 16.5 for deeper (type III) resections, and 15.8 for type IV or V resections. Roh and colleagues observed the same difference and demonstrated that VHI was also worse if the anterior commissure was resected. Thus, patients generally report low levels of voice handicap after TLM with scores comparable or better than voice handicap after treatment for benign laryngeal diseases.

General quality-of-life questions and questionnaires are also used to evaluate patients with early glottic cancer treated with TLM. On a 0 to 100 visual analog scale of patient self-reported impact on everyday life, Aaltonen and colleagues found scores of 31/100 at 6 months and 32/100 at 12 months. Using a Washington University quality-of-life questionnaire and the short version of the Short Form-12, Vilaseca and colleagues were able to show that overall quality of life was close to normal on average (1139 of a total score of 1200 for the Washington University questionnaire) for T1 or T2 glottic cancers, with normal scores in all of the domains except speech. At 1 year, 69% of the patients in their group of glottic and supraglottic cancers had normal Short Form scores, with supraglottic tumors, radiation therapy (RT) and neck dissection having a negative impact on quality of life. The same group found that “negative changes in daily life” were more frequently observed in younger patients than in older patients. Roh and colleagues compared the speech and social contact domains of the European Organisation for Research and Treatment of Cancer head and neck specific module (EORTC QLQ H&N-35) between superficial and deeper glottic resections and found significantly better scores for type I and II cordectomies versus types III and IV. Thus, again, general quality of life seems to be related to the extent of resection and to patient age, but the impact is minimal to moderate at most.