Neoglottal revisions after supracricoid laryngectomy with cricohyoidoepiglottopexy

Supracricoid laryngectomy with cricohyoidoepiglottopexy (SCL-CHEP) is a functional larynx-preserving surgery designed for T2 and well-selected T3–T4 glottic type laryngeal cancers (1–3). By removing ~3/4 of the larynx including thyroid cartilage and the bilateral vocal folds except for one or two arytenoids, SCL-CHEP is capable of achieving three major goals of larynx-preserving surgery: (i) stable oncologic outcomes, (ii) life without tracheostoma and (iii) acceptable vocal and swallowing functions. In the post-SCL-CHEP neoglottis, laryngeal functions are supported by interaction between the remaining arytenoid and epiglottis; the cricoarytenoid unit is known to play a key role in post-SCL functions (4). To obtain optimal functional results, proper surgical techniques must be used (5). Regardless of operating with skilled techniques, various post-operative complications may be encountered (6). The early detection and management of such post-operative complications are crucial (7–12). In this study, we analyzed post-SCL-CHEP patients who received secondary surgical revisions because of neoglottal problems. We retrospectively reviewed the clinical features, risk factors and the secondary revisions carried out. We analyzed the patients with laryngeal cancers (squamous cell carcinoma) who received SCL-CHEP between 1997 and 2015 at our institute. We obtained the clinical data from the post-SCL-CHEP patients who required secondary surgical revisions to correct the neoglottal problems. We reviewed: (i) clinical features, (ii) risk factors (radiation history, infection status and arytenoid status) and (iii) secondary revisions carried out. During the impaction pexis process of SCL-CHEP, it is known to be critical to properly align the anterior borders of the hyoid bone and cricoid cartilage for an ideal neoglottis (8). Consequently, we measured the morphologic positions of the hyoid bone and cricoid cartilage using larynx-targeted CT images from the post-SCL-CHEP patients and compared the results among patients with and without secondary revision. A 16-channel multidetector-row CT (Light Speed Ultra 16; GE Healthcare, UK) was utilized for evaluation. CT scanning was performed under the following conditions: (i) X-ray tube voltage 120 kV, (ii) X-ray tube current 250 mA, (iii) slice thickness 0.125 mm and (iv) slice time 0.6 seconds. Sagittal cut CT images at the mid-line were utilized for the morphologic measurement. To exclude the early stage morphologic instability, only the patients with CT images 1 year after SCL-CHEP were included for the measurement. Morphologic measurement involved: (i) identifying the anterior arches of the hyoid bone (A: lower edge) and cricoid cartilage (B: upper edge), (ii) drawing a straight line from A to vertically reach the anterior border of the cervical vertebrae (C) and (iii) similarly drawing from B to the vertebral border (D). The distances between A and B (cricohyoid gap), A and C (position of hyoid) and B and D (position of cricoid) were measured (Fig. 1). Among the 119 patients, by excluding those with different CT scanning conditions, follow up <1 year, and lost to follow up, data were available for those with and without secondary revision in 13 and 46 patients, respectively. Because of the selected number of the patients included in the morphologic measurements, there may be a limitation to the interpretation of the results. Morphologic measurements of post-supracricoid laryngectomy with cricohyoidoepiglottopexy (SCL-CHEP) larynx with definitions of points A, B, C and D. In the patients with revision (rev+), the cricoid position (B–D) tended to be longer than the hyoid position (A–C), theoretically creating a neoglottal stricture. Among the 119 patients who received SCL-CHEP between 1997 and 2015, 24 patients (20.1%) underwent secondary revisions to correct the neoglottal problems. Pathological TN (pT, pN) staging of the 119 patients analyzed are presented in Table 1 with numbers in parentheses indicate the pTN staging of the 24 patients who underwent secondary revision. There were 23 male and 1 female patient, with a median age of 62.4 years old (ages ranged from 32 to 71 years old, at the time of SCL-CHEP). Pathological TN (pT, pN) staging of the 119 patients analyzed (numbers in parentheses indicate the pTN staging of the 24 patients who underwent secondary revision) Pathological TN (pT, pN) staging of the 119 patients analyzed (numbers in parentheses indicate the pTN staging of the 24 patients who underwent secondary revision) Clinical details of the neoglottal status, etiology and revisions carried out are presented in Table 2. Among the 24 patients, 14 (58%) demonstrated neoglottal stricture due to the redundant mucosa of the arytenoid and/or granulation, and were operated on with endoscopic resection. Four patients showed enlargement of the neoglottis, and buccal fat augmentation was performed. Four patients showed an inverted epiglottis and were treated by epiglottectomy through an endoscopic approach. Two patients exhibited cricoid chondritis due to cervical spur irritation, and they were treated by spur resection through an open approach. Clinical details of the neoglottic status, etiology and revisions carried out Clinical details of the neoglottic status, etiology and revisions carried out Clinical details of morphologic measurements are demonstrated in Table 3. The average A–B (cricohyoid gap) values were 11.4 and 10.9 mm in the patients with and without intervention, respectively. The average A–C (position of hyoid) values were 23.7 and 22.4 mm in the patients with and without revision, respectively. The average B–D (position of cricoid) values were 24.2 and 21.8 mm in the patients with and without revision, respectively. A–C was longer or equal to B–D in 7 and 41 patients with and without intervention, respectively; A–C was shorter than B–D in 6 and 5 patients with and without revision, respectively. The χ2 test revealed a significant difference between patient groups (P = 0.009). Results of morphologic measurements and analyses of related risk factors Asterisk denotes P value being statistically significant. Results of morphologic measurements and analyses of related risk factors Asterisk denotes P value being statistically significant. Results of analyses using other risk factors including a history of radiation, post-SCL-CHEP infection and the arytenoid status are indicated in the lower half of Table 3. Only the post-SCL-CHEP arytenoid status revealed a statistically significant difference between patient groups with and without revision. The tracheal stoma was successfully closed in all 24 patients after secondary revisions. The average duration until surgical closure was 5.7 (18 cases) and 1.4 (70 cases) months for the patients with and without secondary revisions, respectively. The analysis of average duration was limited to those who received surgical closure of the tracheal stoma. SCL-CHEP is one of the open laryngeal preservation surgeries with consistently reliable oncological and functional outcomes. The surgical techniques are characterized by a radical resection of the larynx combined with a unique reconstruction process utilizing three-string impaction. As SCL-CHEP gains increasing acceptance, it is inevitable that surgical complications may be increasingly encountered (6). Among the post-SCL-CHEP complications, laryngeal stenosis has been reported to occur at a rate of 3.7% (9). Pathogeneses in early-onset patients have been reported to be relapse of epiglottis, synechiae, a ruptured pexis and excessive granulation; for late-onset patients, the pathogenesis was related to an excessive healing process at the cricohyoidal impaction. In our series, 24 of 119 (20.1%) post-SCL-CHEP patients underwent secondary revisions. The most frequent pathogenesis was stricture of the neoglottis (20/24 = 83.3%), followed by enlargement of the neoglottis (4/24 = 16.7%). Redundant mucosa and/or granulation at the neoglottis, an inverted epiglottis and cricoid chondritis were the three major causes of neoglottal stricture. Redundant mucosa at the neoglottis was previously reported as ‘flaccid neoglottis’ (11). The main pathogeneses were excessive anterior retraction of the residual mucosa and/or excessive mucosal pliability due to an advanced age. In contrast, excessive granulation was more frequently identified in younger patients. Both pathogeneses (redundant mucosa and granulation) can be easily managed by endoscopic resection of the responsible sites. However, caution needs to be taken to avoid over-resection of the neoglottal inlet, which may deteriorate the laryngeal function. In patients with an inverted epiglottis, the tip of the epiglottis prolapsed and fitted into the neoglottis (10). Technical errors at SCL-CHEP, such as excessive resection of the epiglottic petiole and/or incorrect sutures at the epiglottis, are known to be the main pathogeneses. An inverted epiglottis can be managed by endoscopic resection of the remaining epiglottis, and over-resection needs to be avoided here as well. Cricoid chondritis due to a cervical spur is a rare complication leading to neoglottal stricture. This complication is mostly caused by a C3-4 cervical spur physically contacting the elevated cricoid cartilage (7). When the cervical spur is located at C5-7, which is the common location for Forestier disease, the risk of collision is limited. Resection of the responsible spur through an open surgical approach is required and laryngeal functions can be recovered. Enlargement of the neoglottis is a complication much more difficult to manage than stricture. In our series, four patients (4/24 = 16.7%) were treated with buccal fat augmentation. Autologous buccal fat was harvested transorally and injected endoscopically into the widest plane of the neoglottal inlet (12). Among the four patients, two showed a tightened neoglottis that persisted during observation for more than 24 months (Fig. 2); these patients experienced improved vocal and swallowing functions. However, in two patients, the fat bulge was quickly absorbed after a few months. Fat injection, if conducted properly and the bulge is not absorbed, is a promising approach for managing enlargement of the neoglottis. Sagittal cut CT scans at the mid-line and neoglottal positions on phonation and inspiration in a patient with autologous fat augmentation (a: pre-augmentation, b: 1 month after augmentation and c: 24 months after augmentation). The neoglottis was tightened shortly after augmentation and persisted for 24 months of observation. Morphological measurement revealed that the distances of the cricoarytenoid gap, position of the hyoid and position of the cricoid did not vary among the patients with or without revision. However, when the position of cricoid is longer than that of the hyoid, patients are susceptible to neoglottal stricture, resulting in secondary revision (Fig. 1). This result can be theoretically understood as when the position of the hyoid is shorter than that of the cricoid, the neoglottis can become compressed, resulting in a neoglottal stricture (Fig. 1 rev+). As described by Laccourreye et al. (8), proper and tight alignment of the anterior borders of the hyoid and cricoid is key to avoid this complication. Among the other risk factors analyzed, only the post-SCL-CHEP arytenoid status revealed a statistically significant difference between patient groups with and without revision. When one arytenoid is resected during SCL-CHEP procedure, it is necessary to retract the remaining ipsilateral corniculate cartilage forward to tighten the neoglottis. By doing so, the neoglottis is known to be susceptible to stricture than the patients with bilateral arytenoid status. To diminish the risk of secondary revision, over retraction of the remaining ipsilateral corniculate cartilage should be avoided. In conclusion, ~20% of the post-SCL-CHEP patients required secondary revisions to achieve a favorable neoglottis. After secondary revision, the tracheal stoma was successfully closed in all 24 patients. Head and neck surgeons should be aware of the high risk of neoglottal complications after SCL-CHEP, and so need to be well prepared for secondary revisions. By promptly detecting and adopting proper revisions, neoglottal complications can be corrected and laryngeal functions recovered. This study was supported by a Grant-in-Aid for Scientific Research (C) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (#.24592608: 2012-2016). None declared.