Speech Rehabilitation with Free Lateral Upper Arm Flap after Total or Subtotal Laryngectomy in Patients with Advanced Laryngeal Cancer

Abstract

Background:

Restoration of speech after total laryngectomy (TLG) remains challenging. This study evaluated the feasibility of free lateral upper arm flap (LUAF) reconstruction of a laryngeal phonatory tube and compared postoperative voice outcomes with esophageal speech (ES).

Methods:

The authors retrospectively reviewed 74 men with advanced laryngeal cancer treated between 2019 and 2024. Twenty-three patients underwent LUAF reconstruction of a phonatory tube following TLG or subtotal laryngectomy, while 51 patients received postoperative ES training. Voice outcomes at 3 months were assessed using subjective ratings, objective acoustic measures, and voice-related quality-of-life questionnaires. The mean follow-up time was 28 months.

Results:

Twenty of 23 flaps survived uneventfully, and 3 patients developed transient pharyngeal fistulas. All patients in the LUAF group regained phonation, whereas 26 of 51 patients achieved ES. Compared with ES, LUAF reconstruction resulted in significantly better fundamental frequency, sound pressure level, maximum phonation time, harmonics-to-noise ratio, speech recognition, and lower speech effort (all P < 0.05). Daily communication satisfaction, speech naturalness, and intelligibility were comparable between groups, but swallowing function was inferior in the LUAF group (P < 0.05). No tumor recurrence was observed in the LUAF cohort during follow-up.

Conclusions:

Free LUAF reconstruction provides reliable restoration of phonation after TLG or subtotal laryngectomy with acceptable donor-site morbidity. Compared with ES, this technique enables more consistent voice acquisition and superior voice quality, representing a feasible reconstructive option for speech rehabilitation in selected patients with advanced laryngeal cancer.

Total laryngectomy (TLG) is a common surgical treatment modality for advanced laryngeal cancer. For more than a century, speech rehabilitation has remained a challenge for both surgeons and speech-language pathologists (SLPs). After TLG, there are 3 main methods to restore speech: esophageal speech (ES), electronic larynx, and tracheoesophageal puncture (TEP) with prosthesis. Each method has advantages and disadvantages. The method used for speech rehabilitation depends on the surgeon’s clinical experience, the patient’s adaptability, or the patient’s financial situation. Patients who are unable to communicate through these methods will encounter significant obstacles in their daily lives after surgery. In addition, laryngeal loss presents a psychological burden to patients, some of whom lose not only their ability to speak but also their willingness to try speech rehabilitation.

The speech rehabilitation methods used after TLG or subtotal laryngectomy (sTLG) vary widely, have progressed substantially, and are under debate. Cricohyoidoepiglottopexy and cricohyoidopexy represent the primary efforts to restore phonation after sTLG in patients with laryngeal cancer.¹

Several successful cases of multiorgan complex transplantation, including the larynx, have been reported, but ethical concerns related to allogeneic laryngeal transplantation remain a major challenge.^2–4 Reconstruction of the laryngeal phonatory tube using free flaps, including forearm or anterolateral thigh (ALT) flaps, have been attempted after TLG, with satisfactory results.^5–8 Normally, the epiglottis plays a key role in the swallowing process, as it prevents aspiration. In many laryngeal cancer cases, the epiglottis is free from tumor and can be spared during the repair of circumferential defects of the larynx and trachea.⁹

The free lateral upper arm flap (LUAF) is a fasciocutaneous flap increasingly used in head and neck reconstruction as it is associated with a low rate of donor-site morbidity. We report our preliminary experience with the repair of laryngotracheal defects using a free LUAF. We also compared the voice assessment results of patients in the LUAF group with those who underwent TLG and acquired ES during the same period.

PATIENTS AND METHODS

Ethics Approval

The study was approved by the Medical Ethics Committee of West China Hospital, Sichuan University, and was conducted in compliance with the Declaration of Helsinki and its amendments. Patients provided written informed consent and agreed to attend follow-up examinations at all scheduled times.

Patient Data and Surgical Techniques

A total of 74 patients who underwent TLG or sTLG in our hospital from 2019 through 2024 were retrospectively reviewed. TLG is defined as the removal of any part of the anatomical larynx. sTLG is defined as the removal of all laryngeal structures except for the suprahyoid epiglottis. Exclusion criteria were lack of postoperative speech rehabilitation training, speech using an electronic larynx or TEP, no LUAF, age older than 75 years, or serious underlying disease.

Among the included patients, 23 underwent laryngeal reconstruction using a free LUAF. The remaining patients (51 of 74) received ES training under the guidance of a professional SLP 20 days after surgery. Only 26 patients (26 of 51) in the ES group developed esophageal phonation; all 23 patients in the LUAF group gained voice function after surgery. Postoperative speech therapy was scheduled once per week, with each session lasting 30 minutes. The initial phase of training was planned to last for 2 to 3 months, with a total duration of approximately 4 hours. The training was performed by one SLP in the hospital, who has expertise in the field of voice training and assessment.¹⁰ The voice assessments (3 months after the operation) were performed on 49 patients (of 74) who acquired phonatory function within 1 to 3 months after surgery. Both groups of patients exhibited excellent adherence.

All patients had been diagnosed with advanced laryngeal cancer T4N1-3M0, stage 4 (per 2017 American Joint Committee on Cancer guidelines [AJCC]). The patients were all men, 51 to 75 years of age (median, 60 years). In the LUAF group (n = 23), 14 patients underwent TLG and 9 underwent sTLG in which the partial suprahyoid epiglottis was preserved with negative frozen-section margins. A free LUAF was used to repair the circumferential defect of the larynx and trachea, creating a laryngotracheal tube for the convenience of speech rehabilitation. The size of the flap ranged from 5 × 6 cm to 6 × 9 cm.

Patients were encouraged to speak on the 10th to 14th day after surgery based on the condition of the wound. The appearance and function of the donor site were also evaluated. The TLG without free flap reconstructive surgery group (ES group) received ES training. All patients underwent tracheotomy and bilateral neck dissection simultaneously. A total of 18 LUAF patients (of 23) and 20 ES patients (of 26) received adjuvant radiochemotherapy after surgery, and the remaining 11 patients (of 49) received surgical treatment only. The follow-up time ranged from 1 to 4 years (average, 2 years and 4 months). The evaluators of speech assessment and data analysts were blinded to group allocation.

Speech Assessments

Subjective speech assessments including subjective voice evaluation and voice-related quality of life (VRQoL) assessment using designed 5-point Likert scale questionnaires were performed. The questionnaire items were chosen based on previous literature and our clinical experience.

Patients were asked to produce a complete sentence containing more than 10 characters and fewer than 25 characters in Mandarin. Subjective voice evaluation included daily communication satisfaction (DCS) (5 = satisfactory, 1 = unsatisfactory), articulation effort (5 = effortless, 1= difficult), intelligibility (5 = intelligible, 1 = unintelligible), and naturalness (5 = close to normal, 1 = unnatural). The first 2 items were assessed by the patients and the latter 2 were evaluated by 2 SLPs based on patients’ recorded audio samples. The average of their scores was taken as the result.

VRQoL assessment included social activity engagement (5 = not at all involved, 1 = very involved), swallowing function (5 = able to drink all at once without choking, 1 = multiple episodes of choking and difficulty completing drinking), and dependence on pronunciation aids (5 = independent, 1 = completely dependent).

Swallowing function was evaluated by the Kubota Water Swallow Test (WST). The patients were asked to drink 30 mL of warm water in a sitting or semireclining position, and manifestations during drinking, such as bucking and number of swallows, were recorded.

Study-specific 5-point Likert scales were adapted from previous work and practice (not fully validated instruments [eg, Voice-Related Quality of Life, Voice Handicap Index–10, European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire–Head and Neck Cancer]).¹¹

The ATMOS voice assessment system (product model LingWAVES) and Mandarin speech recognition software (IBM ViaVoice 9.1) were used for objective voice evaluation. Fundamental frequency (F0), sound pressure level (SPL), maximum phonation time (MPT), harmonics-to-noise ratio (HNR), and speech recognition rate (SRR) were recorded. Based on the SRR scores, voice intelligibility was classified into 5 levels: 1 point, recognition rate less than 20%; 2 points, 20% to 40% recognition rate; 3 points, 40% to 60% recognition rate; 4 points, 60% to 80% recognition rate; 5 points, recognition rate greater than 80%.

Statistical Analysis

All statistical analyses were performed using SPSS software (version 22.0; IBM). We compiled the total score for each test. When a normal distribution was present, comparisons between the 2 groups were conducted using independent samples t tests, and reported as the mean ± SD. Otherwise, nonparametric test methods, such as the Mann-Whitney-Wilcoxon test, were used, and the results were reported as median (lower quartile, upper quartile). The chi-square test was used to analyze the difference in rates between 2 groups. P values less than 0.05 were considered significant. Effect sizes (Hedges r) and 95% CIs were reported. Post hoc power analysis was conducted to assess detectable effect sizes and overall study power, while acknowledging potential limitations due to the small sample size and retrospective study design.

RESULTS

In the LUAF group (n = 23), the flaps survived without complications in 20 patients and transient pharyngeal fistulas developed in 3. Primary healing of the upper arm donor sites was also achieved. At the end of the follow-up period, all 23 patients were alive and no signs of recurrence were observed.

Laryngoscopy revealed that the reconstructed laryngeal tube entrance was slit-shaped. The airway below the entrance gradually increased in size and connected to the trachea. One month after surgery, the gastric tube was withdrawn and oral feeding was allowed. Seven patients experienced aspiration, but none of the 23 patients experienced aspiration pneumonia. None of the 23 patients had been extubated at the end of the follow-up. Eighteen patients were able to block the tracheotomy opening for 10 to 30 minutes intermittently, and 5 were able to block the opening for 2 to 10 minutes. Because the flap epithelium could not expel phlegm from the respiratory tract or perform self-cleaning functions, the patients were encouraged to block the tube intermittently to facilitate the elimination of phlegm scabs or epithelial debris from within the flap tube. All patients were satisfied with the postoperative neck contour.

No statistically significant differences were observed between the 2 groups in age, educational attainment, Charlson Comorbidity Index score, cancer subtype, or tumor stage (Table 1). The subjective voice questionnaire scores (Table 2) showed that the patients and SLPs subjectively believed voice quality was comparable between the ES and LUAF groups in the 3 aspects of DCS, intelligibility, and naturalness, indicating a similar voice handicap in the 2 groups (P > 0.05). However, phonation was significantly more laborious in the ES group than in the LAUF group (z = −2.241; P < 0.05).

Table 1.

Patient Demographics^a

Characteristics	ES Group	LUAF Group	Statistics	P
Patients, n	26	23
Age, yrs	59.96 ± 6.056	62.08 ± 7.494	t = 1.097	0.278
1, 2, 3 LNs, %b	53.8, 30.8, 15.4	65.2, 21.7, 13.0	χ2 = 0.689	0.709
Laryngeal cancer subtype
Glottic	22 (84.6)	19 (82.6)	χ2 = 0.571	0.752
Supraglottic	3 (11.5)	2 (8.7)
Subglottic	1 (3.8)	2 (8.7)
Education level			χ2 = 0.119	0.989
Primary school	3 (11.5)	3 (13.0)
Junior high	10 (38.5)	9 (39.1)
Senior high	9 (34.6)	7 (30.4)
University	4 (15.4)	4 (17.4)
Charlson Comorbidity Index	2.00 ± 0.69	2.00 ± 0.85	t = 0.000	1.000

^aData are presented as n (%) or mean ± SD.

^bInvolvement of regional lymph nodes (LNs). Staging ranges from N1 to N3 and reflects the severity of lymph node metastasis.

Table 2.

Comparison of Subjective Voice Analysis Measures between the ES and LUAF Groups^a

Measures	ES Group	LUAF Group	z Statistics	P	Effect Size r	Approximate Achieved Power
DCS	4.00 (3.00, 4.00)	3.00 (3.00, 4.00)	−0.825	0.409	0.12	0.14
Articulation effort	3.00 (2.00, 3.25)	4.00 (3.00, 4.00)	−2.241	0.025	0.32	0.53
Intelligibility	3.00 (3.00, 4.00)	4.00 (3.00, 5.00)	−1.090	0.276	0.16	0.24
Naturalness	3.00 (2.00, 4.00)	4.00 (3.00, 4.00)	−2.265	0.023	0.32	0.53

^aEffect sizes for nonparametric tests were calculated as r = Z /√N. Achieved power was computed post hoc for 2-tailed α = 0.05. With n = 23 versus 26 and α = 0.05 (2-tailed), the study has approximately 80% power to detect r = 0.39 or greater (approximate d = 0.84 or greater). Smaller effects may remain undetected due to limited power.

VRQoL scores (Table 3) revealed no statistical differences between the 2 groups in the degree of social participation or dependence on pronunciation aids. Because the ES group had undergone permanent tracheostomy, the swallowing function of the control group was significantly better than that of the LAUF group (z = −5.663; P < 0.05), and there was no bucking in the ES group. VRQoL effect size (r = 0.23) was below the study’s detectable threshold (r ≥ 0.39), suggesting that nonsignificance may reflect limited power, not true absence of difference.

Table 3.

Comparison of VRQoL Measures between the ES and LUAF Groups^a

Measures	ES Group	LUAF Group	z Score	P	Effect Size r	Approximate Achieved Power
Social activity engagement	4.00 (3.00, 4.00)	3.00 (3.00, 4.00)	−1.626	0.104	0.23	0.35
Swallowing function	5.00 (5.00, 5.00)	3.00 (2.00, 4.00)	−5.663	0.000	0.81	>0.99
Dependence on pronunciation aids	3.5 (3.00, 4.25)	4.00 (3.00, 5.00)	−1.626	0.104	0.23	0.35

^aEffect sizes for nonparametric tests were calculated as r = Z /√N. Achieved power was computed post hoc for 2-tailed α = 0.05. With n = 23 versus 26 and α = 0.05 (2-tailed), the study has approximately 80% power to detect r = 0.39 or greater (approximate d = 0.84 or greater). Smaller effects may remain undetected due to limited power.

We evaluated the objective voice quality in terms of 4 parameters: F0, SPL, MPT, and HNR (Table 4). The LUAF group achieved better scores for all these measures (P = 0.002, 0.002, 0.000, and 0.030, respectively). Speech recognition accuracy, as measured by software, was also significantly higher in the LUAF group compared with the ES group (z = −2.050; P < 0.05).

Table 4.

Comparison of Objective Voice Analysis Measures between the ES and LUAF Groups^a

Measures	ES Group	LUAF Group	Statistics	P	Hedges g (95% CI)	Approximate Achieved Power
F0, Hz	75.1 ± 12.404	86.26 ± 11.046	t = 3.303	0.002	0.94 (0.35–1.53)	0.86
SPL, dB	63.38 ± 8.471	72.00 ± 9.605	t = 3.337	0.002	0.94 (0.35–1.54)	0.86
MPT, s	3.96 ± 1.948	10.52 ± 3.072	t = 8.793	0.000	2.58 (1.81–3.35)	>0.99
HNR, dB	6.57 ± 2.548	8.26 ± 2.733	t = 2.231	0.030	0.64 (0.06–1.22)	0.53
SRR	3.00 (2.00, 4.00)	4.00 (3.00, 4.00)	z = −2.050	0.040	r = 0.29	0.47

^aEffect sizes for continuous variables are Hedges g with 95% CI; for nonparametric tests, r = Z /√N. Achieved power was computed post hoc for 2-tailed α = 0.05.

TLG without Epiglottic Spare: Typical Case 1

A 59-year-old man had a 1-year history of hoarseness. Laryngoscopy revealed a supraglottic tumor and the biopsy confirmed squamous cell carcinoma (TNM stage T4N1M0 [AJCC]). The patient underwent tracheotomy, bilateral neck dissection, and TLG with frozen safe margin. A 5 × 6.5–cm flap was harvested on the lateral upper arm, shaped into a tube, and connected to the trachea below and the tongue base above (Fig. 1). A tracheostomy tube with cuff was used to prevent aspiration in the early postoperative course. A silicone nasopharyngeal airway served as a stent in the tube and was removed 10 days later. After surgery, the patient could speak with a clear, audible, and unforced voice continuously for about half an hour before experiencing mild phonatory fatigue. Pathology revealed metastasis in 1 of 31 left cervical lymph nodes and no metastasis in 34 right cervical lymph nodes. The patient received adjuvant radiochemotherapy.

Fig. 1.

Typical case 1: preoperative enhanced CT of the neck and flap recipient and donor areas. (Above, left) CT scan revealed a tumor (asterisk) with larynx involvement. (Above, right) Wound after TLG and bilateral neck dissection. (Below, left) LUAF harvest: right LUAF with thick subcutaneous fat. (Below, right) The flap was tubulated with a laryngeal stent. F, LUAF; PP, posterior wall of pharynx; RN, radial nerve; SCM, sternocleidomastoid muscle; TB, tongue base; TM, triceps muscle.

Three months after surgery, acoustic voice analysis had the following results: F0, 104 Hz; SPL, 90 dB; MPT, 10 s; HNR, 15 dB; and SRR, 80%. The patient reported satisfaction with the results (DCS, 5 points; articulation effort, 4 points), and the SLP rated speech intelligibility as high (intelligibility, 5 points; naturalness, 3 points). Although the gastric tube had been removed, the patient experienced choking with liquids (WST grade 3, score 3).

One year after surgery, laryngoscopy revealed a stable slit-like neoglottis, which remained unchanged over the next 2 years. The flap below the entrance was smooth, and no abnormal vascular configurations were observed on narrow band imaging (Fig. 2). The patient experienced occasional mild aspiration of saliva or liquids but maintained normal oral intake.

Fig. 2.

Typical case 1: 1-year postoperative follow-up of the laryngeal flap tube. (Above, left) Laryngoscopy revealed that the neoglottic entrance (E) was slit-shaped. (Above, right) CT corresponding to that shown in Figure 1 reveals a reduction in the size of the neoglottic entrance and a wide hypopharynx. (Below, left) The flap tube (FT) was slit-shaped and the size of the lumen increased as it approached the trachea. (Below, right) The connection between the lower end of the flap (F) and the trachea (T) stump. The black arrow indicates the tracheostomy opening, with the tracheotomy tube transiently removed for inspection. E, entrance of the larynx; PP, posterior pharyngeal wall; PS, piriform sinus.

sTLG with Preservation of the Suprahyoid Epiglottis: Typical Case 2

A 51-year-old patient had a 6-month history of hoarseness. Laryngoscopy revealed a tumor with partial laryngeal surface of the epiglottis involvement. The biopsy suggested squamous cell carcinoma. The TNM stage was T4N1M0 (AJCC). The patient underwent surgery similar to that reported for typical case 1 except that part of the suprahyoid epiglottis was spared. A 6 × 9–cm LUAF was fashioned and connected to the tongue base and the free margin of the suprahyoid epiglottis above (Fig. 3). No cervical lymphatic metastasis was noted.

Fig. 3.

Typical case 2: preoperative CT and operative field. (Above, left) CT revealed thickened vocal cords and a subglottic neoplasm. (Above, right) Harvest of the LUAF (F). (Below, left) The tubulated flap in place. (Below, right) The operative field at the end of the vascular anastomosis and flap connection. E, suprahyoid epiglottis.

The patient’s voice seemed unaffected by adjuvant radiotherapy. He experienced a transient pharyngeal fistula after oral feeding and recovered after the gastric tube was reinserted for 1 week. Postoperative laryngoscopy revealed a clover-shaped neoglottic entrance. Three-dimensional (3D) reconstruction revealed a cylindrical flap occupying the anatomical location of the original larynx (Fig. 4).

Fig. 4.

Typical case 2: postoperative fiberoptic laryngoscopy examination and 3D reconstruction of the tubulated flap. (Above, left) The clover-shaped neoglottic entrance formed by the tubulated flap. (Above, right) the surface of the tubulated flap lumen between the neoglottic entrance and the lower end of the flap (F) tube. (Below, left) The spacious joint between the trachea (T) and the lower end of the tubulated flap (F). (Below, right) 3D reconstruction of the tubulated flap located in the primary larynx (yellow arrow). E, entrance; PP, posterior pharyngeal wall; PS, piriform sinus.

Three months postoperatively, acoustic voice analysis showed the following results: F0, 95 Hz; SPL, 72 dB; MPT, 15 s; HNR, 9 dB; and SRR, 70%. Subjective assessments were comparable to those reported for typical case 1 (DCS, 4 points; articulation effort, 5 points; intelligibility, 5 points; naturalness, 4 points), with better swallowing function (WST score 5).

DISCUSSION

TLG is the preferred treatment option for advanced laryngeal cancer, whether primary or recurrent, yet speech rehabilitation poses persistent challenges. Cultural beliefs, educational background, and socioeconomic status will inevitably influence treatment choices. Patients who undergo TLG are generally active in postoperative speech rehabilitation, but the results are not always satisfactory.^12,13 In China, the electrolarynx and ES are the primary methods used for voice restoration after TLG.

ES has the advantages of high naturalness, acceptability, and low cost. The main obstacle of practicing ES after TLG is the loss of sound power from the respiratory system, and the need to establish an alternative airflow power from the digestive tract. This shortcoming leads to a long cycle of ES training and a low success rate, especially among older adults or frail patients, as confirmed by our study. In addition, the lack of airflow power in ES often leads to compensatory movements during phonation, which affects breathing–speech coordination.

Apart from ES, postoperative speech rehabilitation using a free flap or infrahyoid myocutaneous flap after TLG has been reported, with the free flap showing superior outcomes compared with free jejunal grafts.^14,15 The reconstruction of the laryngeal airway with an ALT is similar to that of TEP and leads to improvements in the range and clarity of speech, indicating that free flaps are good alternatives for speech rehabilitation after TLG.⁸ Microsurgical technique improvement provides a high success rate for the restoration of phonatory function in patients who undergo sTLG or TLG.

Both forearm and ALT flaps can be used as donor sites for this approach after TLG. The LUAF is a type of fasciocutaneous flap without muscular involvement, so its harvest will cause no muscle-related movement disorders. Coincidentally, most older adult patients with laryngeal cancer have redundant skin or so-called butterfly sleeves, which can meet the flap size requirement. The pedicle of the flap is short but sufficient or even rich for anastomosis in the larynx area because the routine recipient vessel is close to the larynx. Compared with both the forearm flap and ALT, the LUAF is prone to fewer functional impairments and is more aesthetically pleasing.

From this small cohort, we realized that the size of the space formed by the flap is closely related to the postoperative sound volume and may determine whether or how long the tracheotomy opening can be blocked. Second, whether a temporary laryngeal stent is used or not, when the stent is removed, the lumen of the flap tube will eventually collapse due to the lack of tracheal cartilage-like support. This also makes the extubation unsafe and impossible. On the other hand, both of the partial collapse of the flap and the slit-shaped neoglottis can fortunately reduce aspiration. This technology has the potential to become a new method for voice rehabilitation in patients with advanced laryngeal cancer.

The epiglottis is an important anatomical structure for swallowing. Epiglottic carcinoma is a unique type of supraglottic laryngeal carcinoma, as the epiglottis above the hyoid bone is not often involved and can be spared with a safe margin.⁹ Theoretically and practically, an epiglottis in appropriate size precludes postoperative aspiration and vice versa. In this case series, we positioned the laryngeal entrance at the midline, added several more stitches to the flap at the neoglottic entrance in patients who underwent TLG, and used the spared epiglottis to cover the entrance to avoid aspiration. The results of swallowing function evaluation in this study also provide some objective evidence that sTLG with partial epiglottis spared can improve aspiration. However, more cases are needed to strengthen this conclusion. Based on our experience from this case series and other studies, laryngeal entrance coverage deficit is not easy to overcome by functional rehabilitation training or over time. For patients without cervical lymph node metastasis, the value of postoperative radiochemotherapy should be confirmed through controlled studies. The surgical plan also depends on resectability, underlying diseases, and the patient’s financial status.^16,17

This study may be subject to selection bias because the groups have not been randomized. Patients in this study were all men and had a comparable cultural background in terms of occupation and living conditions. The participants had similar tumor staging (T4N1M0 or greater) and basic surgical techniques. Therefore, the baseline characteristics are theoretically comparable. This study is also limited by sample size, with only 23 LUAF cases, creating difficulty in detecting infrequent complications or subtle differences in functional outcomes. In addition, our Likert-based questionnaires were not formally validated instruments; thus, although they captured clinically relevant domains, they may not fully meet psychometric standards. Future studies should incorporate standardized tools, such as the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire–Head and Neck Cancer, global quality of life data, and Voice Handicap Index, to improve generalizability and comparability.^11,18,19

The WST is a simple, inexpensive, and readily accessible screening tool for overt aspiration risk on thin liquids, but its limitations are profound, especially in populations at high risk of silent aspiration. In this study, the WST rather than a standard videofluoroscopic swallow study (VFSS) or fiberoptic endoscopic evaluation (FEES) of swallowing was used because the cause of aspiration was more structural than neurologic. As all patients with LUAF had a functionally intact cough reflex, the critical limitation of WST in identifying silent aspiration was considered less consequential in this context. VFSS and FEES provide accurate diagnosis of silent aspiration, which is beyond the capabilities of the WST. However, these evaluations were not acceptable to most patients. For the purpose of comparison and data collection, standard VFSS or FEES evaluation should be performed in future studies.

CONCLUSIONS

LUAF is a promising candidate to restore phonatory function after TLG or sTLG with acceptable injury to the donor site. Compared with ES, this technology is a feasible alternative for speech rehabilitation. Reducing aspiration and achieving eventual extubation remain key challenges to optimizing this method.

DISCLOSURE

The authors have no financial interests to report.

ACKNOWLEDGMENT

This work was supported by the 1·3·5 project for disciplines of excellence–Clinical Research Fund, West China Hospital, Sichuan University (grant 2024HXFH004).