Functional Outcomes in Partial Glossectomy Defects Reconstructed with and without Free Tissue Transfer

Reconstruction of ablative defects of the tongue pose significant challenges in regards to speech and swallowing outcomes. Although current data generally support the use of free flap reconstruction for hemi-glossectomy defects, question remains regarding optimal reconstruction of defects of less than ½ of the anterior tongue. Reconstructive options for defects which cannot be closed primarily include split-thickness skin grafts, skin graft substitutes (e.g. alloderm, allograft, etc.) , and secondary intention healing. It is unclear what size tongue defect may benefit from free tissue reconstruction vs. the above alternatives for optimal speech and swallowing outcomes. In this study we investigated speech and swallowing outcomes for these patient populations.   

With IRB approval, we prospectively enrolled patients with a diagnosis of squamous cell carcinoma of the tongue who had undergone tongue resections of at least 3.5cm in the greatest dimension (based on gross pathology specimen measurement), had been reconstructed with a free flap, skin graft or alloderm,  and had been followed for at least 1-year post-operatively. Patients who had undergone multiple surgeries of the tongue or who were unwilling to participate by signing consent were excluded. Patient defect size was calculated using gross pathology with a total volumetric (X*Y*Z) number recorded.  All patients completed validated tools to assess swallowing quality of life and speech function namely: Swallowing Quality of Life Assessment Questionnaire (SwolQol); Assessment of Intelligibility of Dysarthric Speech (AIDS).  The SwolQol was completed by all patients and speech samples were obtained by a speech pathologist with an audio recording  which was scored in a blinded fashion by two third party volunteers without formal speech training. Data was analyzed using statistical software with Student’s t-test comparison of groups (flap vs. no flap) and Pearson correlation. A statistical significance of p < 0.05 was considered statistically significant.

In this pilot study, 14 total patients were recruited with 7 patients in each group. There was a statistically significant difference in defect size between the two groups with the Flap group defects five times larger than the Graft group: Flap 56.8cm3 (13.1-154, SD 56.8), Graft 10.09cm3 (3.6-20.4, SD 5.6), P=0.02. There were no differences between word and sentence intelligibility between the two groups: Word intelligibility: Flap 92.14% (88-96%, SD 3.13), Graft 94.4% (87-100%, SD 4.54), p=0.28. Sentence intelligibility: Flap 98.16% (96-100%, SD 1.2), Graft 97.49% (93-100% SD 2.1), p=0.47. There was no statistically significant difference in SwolQol scores: Flap 17.61 (13.7-22, SD 3.25), Graft 19.2 (16.6-20.1, SD 1.79).  

Although only a small number of patients were enrolled at the time of abstract submission, we were able to demonstrate a striking and statistically significant difference in defect size between the two groups. It is notable that intelligibility and swallowing quality of life where similar between the groups and statistically significant differences were not detected. These results should be interpreted with caution given the small numbers but could represent a benefit to patients with larger, but less than hemi-glossectomy, defects of the tongue who function similar to patients with smaller defects that do not undergo flap reconstruction. Flaps may compensate in swallowing quality of life and speech for a five fold increase in defect size based on this very preliminary data.