Abstract
Background/Aim: The tongue is an important anatomical structure, playing an significant role in natural speech, swallowing, and sense of taste. Immediate reconstruction using autologous tissue must be performed following glossectomy for tongue cancer to improve patient quality of life. This study aimed to demonstrate the usefulness of a surgical technique using the free vertical latissimus dorsi flap (FvLDF) for tongue reconstructions using autologous tissue. Patients and Methods: Among patients who underwent total glossectomy for tongue cancer from November 2014 to February 2023, we selected 10 patients who underwent immediate tongue reconstruction with a radial free forearm flap (RFFF) or free anterolateral thigh flap and four patients who underwent FvLDF. The patients were compared regarding postoperative function (width of oropharyngeal space in computed tomography, language-speech evaluation), aesthetic results, and features. Results: All four patients who underwent FvLDF showed successful flap survival, with no severe complications. Because vertical incision was made during flap harvest with primary closure possible with the mid-axillary line, donor morbidity was significantly lower in patients who underwent reconstruction with FvLDF than in those who underwent reconstruction with RFFF, and good aesthetic results were obtained. In comparing the oropharyngeal space of patients on neck CT preoperatively and postoperatively, the width increase rate of patients who underwent reconstruction with FvLDF was significantly smaller. FvLDF patients demonstrated good speech and swallowing functions. Conclusion: Considering the advantages of reconstruction with FvLDF in terms of features and aesthetic results, this surgical technique may be a reliable alternative technique for tongue defects after glossectomy.
- Free vertical latissimus dorsi flap
- glossectomy
- tongue reconstruction
- radial free forearm flap
- tongue cancer
Tongue cancer is a common oral cavity cancer, accounting for 40-50% of oral cavity cancers. The first-choice treatment for tongue cancer is wide excision, i.e., surgical resection, depending on the extent of the tumor. Surgical resection of the tongue is referred to as glossectomy, and it is classified into partial glossectomy, hemi-glossectomy, subtotal glossectomy, and total glossectomy based on the extent of tongue resection. Partial glossectomy removes <1/3 of the tongue; hemiglossectomy removes 1/3-1/2 of the tongue; subtotal glossectomy removes 1/2-2/3 of the tongue; and total glossectomy removes the entire tongue. Because the tongue serves several functions, glossectomy-related lingual defect can cause fatal discomfort in the patient’s quality of life. The tongue not only plays an important role in the patient’s speech, articulation, swallowing, and mastication, but it is also involved in taste, oral hygiene, and prevention of aspiration. Therefore, tongue reconstruction after total glossectomy for tongue cancer is considered an important procedure.
The ideal tongue reconstruction involves complete cover of the lingual defect with a sufficient amount of autologous tissue to restore speech and swallowing function, separation of the mouth and throat to restore the pre-operative state, and prevention of a tongue scar that restricts tongue movement. Therefore, various surgical techniques using autologous tissues are available. Among regional flaps, the infrahyoid fasciomyocutaneous, pedicle submental island, and pectoralis major myocutaneous flap are well-known (1, 2). These regional flaps can be applied only when there is a small defect, such as during a partial glossectomy or hemi-glossectomy. These flaps are not appropriate in total glossectomy, as there is an insufficient amount of tissue to be covered. In addition, they are not applicable if there is a history of extensive neck disease or prior neck surgery or if the pedicle is sacrificed during oncological resection. To overcome the limitations of regional flaps, tongue reconstruction using free flaps was developed (3-5).
As far as is known, for hemi-glossectomy, radial forearm free flap (RFFF) and free anterolateral thigh flap (fALTF) are available options. For subtotal glossectomy, fALTF or free anterolateral thigh myocutaneous flap is recommended. As total glossectomy requires a larger volume than it does for partial glossectomy, free pentagonal anterolateral thigh myocutaneous flap and free latissimus dorsi flap are preferred. Owing to disadvantages of RFFF, such as insufficient amount of tissue, morbidity of the donor site, and poor aesthetic result, fALTF with a sufficient amount of tissue at the donor site and low morbidity have recently been highlighted (3). However, fALTF has the disadvantages that it is suitable only for people with thin skin and that it can cause discomfort to the patient owing to a large volume decrease more than 6 months after surgery. In addition, the volume of the conventional free latissimus dorsi flap is too large to be positioned in the intraoral area, which can be inconvenient for surgery, and the patient may feel uncomfortable after surgery owing to its large volume. Since a position change is essential during the operation, it is inconvenient to perform total glossectomy at the same time. These various factors can contribute to longer operation time.
Therefore, this study aimed to evaluate postoperative speech and swallowing functions in patients who underwent a free vertical latissimus dorsi flap (FvLDF) procedure with tongue shaped design, which can be applied in total glossectomy with a sufficient amount of tissue, leading to good aesthetic results postoperatively, with the donor site scar covered under the arm (6). In addition, speech and swallowing function of the patient group who underwent FvLDF after glossectomy were evaluated and compared with the patient group who underwent RFFF or fALTF, and analyzed.
Patients and Methods
Study design. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The Institutional Review Board of Kyungpook National University Chilgok Hospital (No. 2023-05-014) approved this retrospective study, and all patients provided informed consent to have their data (including de-identified photographs) recorded, analyzed, and published for research purposes.
This study was conducted from November 2014 to February 2023, including a total of 14 patients (control group: 10 RFFF or fALTF patients; experimental group: four FvLDF patients) who underwent total glossectomy and neck dissection for tongue, salivary gland, or floor of mouth cancer. Patient demographics and surgical records were reviewed (Table I).
Patient demographics.
The postoperative evaluation of both groups was compared in terms of flap survival, donor site morbidity, speech, and swallowing function (postoperative day 7, months 1, 6, and 12). We compared swallowing function based on patient’s and clinician’s evaluation as follows: 1) regular diet; 2) soft diet; 3) fluid diet; or 4) nasal feeding. Speech function was evaluated subjectively by the patient, clinician, and a speech therapist in postoperative years 1 and 2. The speech therapist assessed consonant and vowel accuracies at the word and sentence levels of the native language, Korean, respectively.
To objectively evaluate speech and swallowing function, the width of the oropharynx, which serves as the conduit connecting the nasal cavity and oral cavity and allows food to pass into the gastrointestinal (GI) tract during oral intake, was assessed using axial and sagittal views of CT scans. The oropharyngeal width was measured by the long axis length between the tongue base and inferior constrictor muscle in the CT sagittal view of the patient’s neck; in addition, the width was measured using the CT axial view at the same level. F/u neck CT was performed at preoperative and postoperative months 6 and 12, respectively, to confirm the patient’s progress and measure the width of the oropharynx (Figure 1).
A method for measuring the width of the oropharynx space in neck CT. (A) Sagittal view. After measuring the length of the long axis of the oropharynx between the tongue base and inferior constrictor on the upper part of the epiglottis in the sagittal view, the width of the long axis of the oropharynx space was also measured in the axial view at the same level. (B) Axial view. This image is published with the patient’s consent. FvLDF, Free vertical latissimus dorsi flap; CT, computed tomography.
In the present study, patients with underlying neurological disease, GI disease, or impaired speech and swallowing function due to deterioration of mentality were excluded.
RFFF or fALTF patients and technique. All patients underwent upper and lower extremity CT angiographs and Allen test to check vessel patency before surgery. In addition, the dominant arm of each patient was identified, and the opposite arm was used as a donor. A radial forearm flap was designed according to the size of the defect above the cephalic vein, and proximal flap elevation was performed along the outer edge of the designed flap. During flap elevation, dissection was performed up to the superficial surface of the deep fascia, and the subcutaneous tissue between the flap and cephalic vein was preserved as much as possible to protect the communicating branches. The perforators located between the veins were dissected, and the radial artery and cephalic vein were cut off and ligated.
Prior to the fALTF procedure, marking was performed using Doppler ultrasound (US) to find the perforator of the descending branch of the lateral circumflex femoral artery (LCFA), and the design was made according to the size of the skin paddle. During flap elevation, dissection was performed anteriorly to posteriorly direction, paying attention to the perforating vessels of the rectus femoris and vastus lateralis. Afterwards, one or two perforators with good blood flow were found, and the LCFA and vein were cut off and ligated.
After flap harvesting, for the radial artery and cephalic vein in RFFF and LCFA and the vein in fALTF, micro-anastomosis was performed using 8-0 or 9-0 ethilon with one artery and one vein or one artery and two veins in the facial artery, external jugular vein, internal jugular vein, or facial vein, respectively. Thereafter, for tongue defect after glossectomy, flap insetting was performed, and smooth flap circulation was confirmed using Doppler US during surgery.
At the donor site of the RFFF, a split-thickness skin graft was performed using the ipsilateral thigh as a donor, followed by tie-over dressing; a negative drain was applied to the donor site of fALTF and primary closure was performed.
FvLDF patients and technique. All four patients underwent chest CT angiography as a preoperative examination to confirm the patency of the thoracodorsal vessels to be used as pedicles for FvLDF. Among the left and right pedicles, the one with better vessel patency was selected, and the design was constructed before surgery. After marking the anterior and posterior borders of the arm in the upright position, a modified vertical LD flap (mustache shape) was designed to meet the volume and skin paddle required for the mid-axillary line that can be covered by the arm (Figure 2 and Figure 3A). A pinch test was conducted to design a skin paddle sufficient to provide primary closure of the donor site, and the pedicle was confirmed with Doppler US.
Illustration of tongue reconstruction using FvLDF techniques. After designing the flap so that the scar could be covered by the arm in the upright position, the mustache-shaped flap was elevated while maintaining the supine posture intraoperatively. Both ends of the elevated LD flap were joined with 2-0 vicryl to form a tongue shape, and the remaining parts were skin sutured with 2-0 vicryl to minimize tension, resulting in a convex tongue. FvLDF, Free vertical latissimus dorsi flap.
Intraoperative findings of FvLDF. (A) FvLDF design; (B) elevated FvLDF; (C) immediate postoperative findings; (D) one month postoperatively. This image is published with the patient’s consent. FvLDF, Free vertical latissimus dorsi flap.
After glossectomy, appropriate recipient vessels in the head and neck areas were identified using a microscope and arranged; thereafter, LD flap elevation was performed. After injecting 10-15 cc of 1% lidocaine mixed with epinephrine at the subcutaneous level, an incision was made with a 10-blade, following the preoperative design (mustache shape). Flap elevation and bleeding control were performed using the Bovie device, and tagging sutures were performed on the skin and anterior border of the LD muscle to maintain a small perforator. Afterwards, undermining was performed in the axilla direction to find the thoracodorsal artery pedicle, and the pedicle direction was marked using a visually visible part and confirmed with Doppler US. Dissection was made in the inferior direction, and after confirming the main pedicle, humeral dissection and thoracodorsal nerve sacrifice were performed using a Ligasure device. After pinpoint bleeding of the distal margin of the flap was confirmed, a FvLDF was elevated (Figure 3B). For the donor site, 4 cc of fibrin glue was applied evenly, and primary closure was performed after an 800 cc suction drain was installed. For the donor vessel, the facial artery and either the external jugular vein or internal jugular vein were used, with options for either one artery and one vein or one artery and two veins. Micro-anastomosis was performed using the thoracodorsal artery, vein, and 8-0 or 9-0 ethilon sutures. After micro-anastomosis, flap circulation was confirmed using Doppler US and a convex tongue was constructed. Both ends of the elevated LD flap were joined with 2-0 vicryl and sutured to form a tongue. To adequately prevent the application of tension to the rest of the area, subcutaneous layer suture was not performed, and only skin suture was performed with 2-0 vicryl to achieve a convex tongue (Figure 3C).
Statistical analysis. Data analysis and statistics were performed with the help of a professional statistician. All analyses were performed using IBM SPSS Statistics for Windows version 23.0 (IBM Corp., Armonk, NY, USA). Assessment data were determined to be significant according to a homogeneity of variance t-test between the flap groups at each time point (p<0.05).
Results
The average age of the four patients who underwent FvLDF was 47.5 years (range=20-58 years) and average BMI was 21.65; all patients were within the normal weight range. Three patients were diagnosed with tongue cancer, and total glossectomy was performed. The histological diagnosis was confirmed as squamous cell carcinoma in all three patients, and the stages were confirmed as pT3N2bM0, pT4aN2bM0, and pT4aN0M0. One patient was diagnosed with adenoid cystic carcinoma of the salivary gland (stage pT3N0M0), and total glossectomy was performed. The skin paddle size of the experimental group was 97.2±29.4 cm2. In the control group, adjuvant radiotherapy and chemotherapy were performed 0 or 27-33 times and 0 or 3-5 times, respectively. In the experimental group, radiotherapy and chemotherapy were performed 0-30 times and 0-3 times, respectively (Table II).
Functional and subjective evaluations of FvLDF reconstruction.
All four patients who underwent tongue reconstruction using FvLDF confirmed that the flap condition was good without major complications; in addition, the donor site healed without major complications. In addition, on postoperative day 7, and months 1, 6, and 12, the subjective speech function was confirmed as normal in three (75%) patients and slightly impaired in one (25%), and swallowing function was confirmed as regular diet in one (25%) and soft diet in three (75%), indicating that all patients demonstrated good speech and swallowing functions (Figure 4, Table II).
Photo of case examples of patients who underwent tongue reconstruction using FvLDF. (A) three weeks postoperatively; (B) three days postoperatively. The flap condition of both patients was good. This image is published with the patients’ consent. FvLDF, Free vertical latissimus dorsi flap.
The average age of the control group was 68.2 years (range=60-80 years) and average BMI was 21.69; all patients were within the normal weight range except for two underweight patients. Except for one patient diagnosed with floor of mouth cancer, all others were diagnosed with tongue cancer. The staging was determined as follows: T2N0M0 in two patients, T2N2M0 in three patients, T3N0M0 in two patients, T3N2M0 in one patient, T4N0M0 in one patient, and T4N2M0 in one patient. Four out of 10 patients had necrosis at the donor site after split-thickness skin graft was performed up to 6-month postoperative (Table I).
A speech therapist objectively evaluated the accuracy of Korean consonants and vowels at the word level and sentence level on 1-year postoperative and 2-year postoperative, respectively. At 1-year postoperative, consonant accuracy in the control group was as 27.91±13.23% at the word level and 23.26±9.2% at the sentence level and vowel accuracy was 43.2±16.34% at both the word and sentence levels. No significant change was observed at 2-year postoperative. However, in the experimental group, consonant accuracy was 56.66±11.36% at the word level and 47.36±15.31% at the sentence level at 1-year postoperative. Vowel accuracy was 53.1±9.36% at the word level and 49.22±10.3% at the sentence level. Speech function, as evaluated by the speech therapist, showed better results in the experimental group.
The f/u neck CT measure of the oropharynx width, compared to the preoperative value, showed that the mean increase rate of width in the control group was 147.1±13.3% in the axial view and 125.9±13.1% in the sagittal view at 6-month postoperative and 164.7±14.8% in the axial view and 161.9±19.3% in the sagittal view at 1-year postoperative. In the experimental group, the mean increase rate of width in the control group was 130.4±9.2% in the axial view and 107.2±4.4% in the sagittal view at 6-month postoperative, and it was 133.6±9.2% in the axial view and 108.1±5.5% in the sagittal view at 1-year postoperative. Comparing the average increase rates at 6-month postoperative and 1-year postoperative of the two groups, the average increase rates in the axial and sagittal views were significantly smaller in the FvLDF patient group (p=0.0457, 0.0025, 0.0181, 0.0001) (Figure 5 and Figure 6).
A graph of changes in oropharynx width over time in a 55-year-old male patient who underwent FvLDF and a 60-year-old female patient who underwent RFFF after total glossectomy due to tongue cancer. (A) Graph of changes in the axial view. FvLDF patients showed an increase rate of 140.2% at POY 1, and RFFF patients showed an increase rate of 187.2%. (B) Graph of changes in the sagittal view. FvLDF patients showed an increase rate of 99.6% at POY1, and RFFF patients showed an increase rate of 188.6%. FvLDF, Free vertical latissimus dorsi flap; RFFF, Radial free forearm flap; preop, preoperative; POM 6, 6-month postoperative; POY 1, 1-year postoperative.
Average rate of increase in oropharynx width over time. As a result of measuring oropharynx width by performing f/u neck CT, the average increase rate of width in the control group compared to preoperative was 147.1±13.3% in the axial view and 125.9±13.1% in the sagittal view at POM 6, and 164.7±14.8% in the axial view and 161.9±19.3% in the sagittal view at POY 1. In the experimental group, the mean increase rate of width in the control group was 130.4±9.2% in the axial view and 107.2±4.4% in the sagittal view at POM 6, and it was 133.6±9.2% in the axial view and 108.1±5.5% in the sagittal view at POY 1. As a result of comparing the average increase rates at POM 6 and POY 1 of the two groups, the average increase rates in the axial and sagittal views were both significantly smaller in the FvLDF patient group (p=0.0457, 0.0025, 0.0181, 0.0001). POM 6, 6-month postoperative; POY 1, 1-year postoperative; FvLDF, Free vertical latissimus dorsi flap.
Discussion
Defects after surgical resection for tongue cancer have a great impact on patient physiological and social functions (7). Unlike reconstruction of postoperative defects in other organs, plastic surgeons not only have to consider the removal of dead space and cosmetic aspects but also the patient’s ability to speak, swallow, and chew in tongue reconstructions, although tasting function cannot be restored by tongue reconstruction following partial or total glossectomy.
Regional and free flaps can be used for tongue reconstruction. However, Jinbing Liu et al. reported that the use of regional flaps could affect the mobility of the reconstructed tongue and postoperative speech by causing gravity-induced drooping of the tongue and pedicle (8). In addition, Ji Y.B. et al. reported that reconstruction using a free flap in the case of subtotal glossectomy (more than half of the tongue volume) improved postoperative speech and swallowing function as well as the patient’s postoperative quality of life (9-11).
Until recently, reconstruction has mainly been performed using RFFF or fALTF, depending on the extent of the defect after glossectomy (12-15). However, because primary closure of the donor site is impossible in RFFF, a split-thickness skin graft must be performed on the donor site. Consequently, morbidity problems can occur due to complications such as hand motion or sensation, its location in an externally exposed area, and tendon exposure. In particular, since it is a very thin flap with limited volume, it can be considered most suitable for reconstruction of the esophagus or pharynx and can be used for defect coverage after total glossectomy. As swelling subsides and atrophy and contracture progress postoperatively, the oropharyngeal space becomes wider than it was preoperatively, which can increase nasalance and risk of aspiration during oral intake. In the case of fALTF, although it is one of the most basic perforator flaps, it requires a learning curve for novice reconstruction surgeons. It can be challenging to use in patients with excessive muscle bulk or excessive hair, and it is generally considered suitable for patients with thin skin, which is a disadvantage. One of the most important observations when using these flaps for tongue reconstruction, particularly in cases of extensive defects such as total glossectomy, is that immediately after surgery, the flap size, swelling, and congestion may lead to the belief that the oral cavity is completely occluded. However, over time, usually within six months after surgery, there is a rapid improvement in swelling accompanied by atrophy and contracture, resulting in a frequent lack of volume. Furthermore, although the reconstructed tongue cannot move in the same way as the original tongue, it relies on the three-dimensional movement of oral structures to facilitate eating and speech (16, 17). When the volume of the flap decreases and the size of the canal increases, there is an increased risk of aspiration and speech difficulties, which significantly impairs communication with the likelihood of problems arising. Therefore, in this study, we have proposed a design for tongue reconstruction using the LD flap, which is one of the flaps that can provide the largest volume in the human body and is easily accessible for reconstruction surgeons with fewer pedicle variants (18). The design is aimed at achieving a tongue shape that facilitates the reconstruction process.
The reconstruction of intraoral defects using RFFF was first introduced in 1983. Subsequently, reconstruction methods using various flaps for intraoral defects have been developed to optimize postoperative function. Among them, the LD flap was first introduced by Iginio Tansini in 1906 and first used for breast reconstruction in the 1970s. Later, in 1978, after Maxwell announced that the free LD flap could be used for the reconstruction of the head neck, and lower limbs, substantial progress was made; nowadays, the free LD flap is used for large defects of the head or large muscle and bone defects of the face (19-21).
Previously, we reported the usefulness of the Boomerang LD flap, a method to increase the volume in total breast reconstruction and increase the projection of the reconstructed breast upper pole by transforming the conventional LD flap, which has an oval shape design, and the pedicled vertical LD flap in partial breast reconstruction (22). Therefore, based on the vertical LD flap, the boomerang LD flap design was merged to make it similar to the tongue shape and facilitate surgery.
In particular, in the vertical LD flap, since position change during surgery is unnecessary, the total operation time is significantly lower than that of the conventional LD flap, and it is useful for the reconstruction of the head and neck. Primary closure is possible, as the donor site does not require secondary surgery (skin graft), and the aesthetic result is excellent, as the linear scar is covered by the arm when it is upright. A convex-shaped tongue is important for the patient’s speech and swallowing functions, and a slightly wider design is needed on the mouth floor. In addition, it is an important part of preventing aspiration during eating by separating the throat and oral cavity. Therefore, we designed a mustache-shaped skin paddle shape instead of the existing oval-shaped skin paddle, and we performed a tongue reconstruction of the ideal convex tongue shape (Figure 2).
Four patients who underwent tongue reconstruction using FvLDF and 10 patients who underwent reconstruction using RFFF or fALTF were retrospectively evaluated. There was no significant difference in flap survival rate, but the appearance of the reconstructed tongue was better in the patient group reconstructed using FvLDF, and the experimental group was superior in cosmetic aspects, including donor site morbidity. In addition, the rate of increase in the oropharyngeal width confirmed using neck CT at preoperative and postoperative months 6 and 12 was significantly lower in the experimental group than in the control group (Figure 6). Because the oropharynx is the first canal that leads from the oral cavity to the GI tract during oral intake, it was believed that the small increase rate of the oropharynx was indicative of a low probability of aspiration if the tongue could not move. In addition, it was considered that nasalance was lowered during vocalization, thereby facilitating communication. In addition, when swallowing level (regular diet, soft diet, fluid diet, or nasal feeding) was evaluated by the patient, the same result was shown. In postoperative years 1 and 2, the results of the evaluation of consonant accuracy and vowel accuracy at the word and sentence levels of Korean by a speech therapist showed that the experimental group had significantly better language test results than did the control group (Table II).
The limitations of this study include a small patient sample size owing to the retrospective nature of the study. Therefore, the size and volume of the tongue could not be prospectively measured before surgery and correlated with the flap design. We plan to conduct future studies on a reconstructed tongue using an LD flap that can move by grafting even motor nerves.
Conclusion
The use of the free vertical latissimus dorsi flap in tongue reconstruction after total glossectomy demonstrated more reliable results in terms of postoperative functions, such as aesthetic result, speech function, and swallowing function, as well as donor morbidity, than did RFFF or fALTF, which were mainly used in the past. In addition, the free vertical latissimus dorsi flap is considered reliable as an alternative surgical technique.
Footnotes
Authors’ Contributions
GilJoon Lee: Conceptualization, data curation, writing; Yong June Chang: Conceptualization, formal analysis, writing; Jongmoo Park: Data curation; Jeong Yeop Ryu: Methodology; Kang Young Choi: Methodology; Jung Dug Yang: Formal analysis; Ho Yun Chung: Formal analysis, data curation; Byung Chae Cho: Conceptualization; Byungju Kang: Methodology, data curation; Jeeyeon Lee: Project administration; Joon Seok Lee: Conceptualization, review & editing.
Conflicts of Interest
The Authors declare no conflicts of interest.
- Received June 17, 2023.
- Revision received August 6, 2023.
- Accepted August 28, 2023.
- Copyright © 2023, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).