2018 Section 6 - Laryngology, Voice Disorders, and Bronchoesophalogy

followed in the ambulatory setting without the need for repeat operative intervention. Yamamoto and colleagues 43 previously attempted to compare treatment options for subglottic stenosis. Their systematic review and meta-analysis showed there were success rates for LTR of 95%, laryngoplasty with or without cartilage grafting of 76%, and endoscopic suc- cess rates between 40% and 82%. This group defined success as no further treatment, and their results were limited to reports with at least 10 cases. They concluded that endoscopic treatments may be indicated for short regions of stenosis under 1 cm, and that cases with glot- tic involvement may be better treated with laryngoplasty with or without cartilage grafting. 43 Our results include additional articles and provide some similar findings, with a different set of statistical analysis. In this series, we examined various causes of steno- sis (trauma, intubation/tracheostomy, or idiopathic) to determine whether surgical management should be influenced by etiology. Patients with trauma as the cause of stenosis were more likely to need further sur- gery compared to those with intubation/tracheostomy or idiopathic stenosis. A possible explanation is that both blunt and penetrating trauma can cause disruption of the laryngotracheal framework. 51 Idiopathic stenosis, by comparison, shows histopathological change to the lam- ina propria, leaving the cartilaginous framework intact. 42,52 Our series also shows that etiology of stenosis can affect decannulation rates. The intubation/tracheos- tomy group showed the highest rate of decannulation. Similar to the soft tissue injuries of idiopathic stenosis, this may be due to limited disruption of the laryngeal framework from the tracheostomy or endotracheal tubes. 51 Though intubation may cause tissue damage in 93% of patients, stenosis only occurs in 6% to 10%, fre- quently due to prolonged intubation, lack of proper seda- tion, or gastroesophageal reflux disease. 1,51,53 Thus, unless the injuries are severe, they usually heal without development of clinically relevant stenosis. Although this review attempted to extract data on grade, length, and location of stenosis, reports were incon- sistent and nonstandardized, as 13 articles discussed grade of stenosis 1–3,6,8,11,13,14,25,27,28,30,31 and six articles discussed length of stenosis. 1,17,24–26,28 Grade of stenosis, often reported as the Cotton-Meyer score, 54 is extrapo- lated from the pediatric literature, and is sporadically reported in cases discussing adult LTS. The concomitant presence of vocal fold immobility or laryngopharyngeal reflux were also not reported in the majority of the articles. Additionally, there was insufficient data to include voice quality as an outcome measure. In this series, as well as in the individual observatio- nal studies included in this series, bias was introduced and may affect outcomes. We used the MINORS criteria to eval- uate bias in each study, and due to their observational nature, conclusions are limited. 47 Several of these series are from tertiary care or referral centers that have exper- tise in these treatments, potentially prohibiting widespread application of the results, limiting external validity. Treat- ment allocation is not standardized in observational stud- ies, and therefore an intervention may be chosen based on

an individual surgeon or patient preference, rather than a risk-severity scale, patient-specific characteristics, or other means of allocation. This precludes conclusions for which treatment is best, as compared to randomized control trials or studies from a prospectively collected database that would be more likely to definitively prove one treatment is better than another. Furthermore, in regard to the outcome measure of receipt of additional surgery, criteria used to perform additional procedures are not explicitly stated in the vast majority of studies. The authors of this study may have included bias into this study as well. We did not include all locations of stenosis, such as isolated glottis or tracheal stenosis. In addition, only papers written in English were included, excluding multiple publications in Chinese that may have affected the outcomes of this study. An individual data meta-analysis, 55 in which the authors of the listed studies are contacted to obtain original data, would also have provided further insight, but due to large sample size and the variable age of the studies, this was not part of the study protocol. Further analysis of the original data could allow formulation of a risk-severity scale. Based on this review, surgical management of air- way stenosis should be tailored to the type of stenosis, both in its severity and etiology. Endoscopic manage- ment should be reserved for less severe cases, or those in which the laryngotracheal framework is intact. Using prospectively acquired data, a treatment algorithm for LTS patients can be designed and further tested. There may be tremendous power from a prospective multi- institutional database that tracks characteristics of ste- nosis such as grade, length, and location; etiologies; patient variables such as pulmonary function, voice quality, quality of life, vocal fold mobility, and presence of reflux; and surgical outcomes including additional surgery, decannulation, and need for endoscopic debride- ment after open surgery. Consistent documentation of grade, length, and location of stenosis may allow formu- lation of a risk-severity scale, whereas inclusion of sur- veillance and follow-up procedures would likely provide important information for patient counseling. Finally, although adjuvant treatments such as mitomycin C and steroids have been described, their role in the treatment of LTS is not well defined, and tracking their use could determine efficacy 1,3,4 CONCLUSION Patients with adult LTS who undergo laryngotra- cheal resection with anastomosis receive less surgery compared to those who undergo endoscopic treatment or LTR with augmentation/grafting. Patients with idio- pathic stenosis are less likely to receive further surgery compared to those from trauma or intubation/tracheos- tomy, but have the lowest rate of decannulation. How- ever, as these results are based on observational studies, rather than studies with randomized or prospective designs, definitive conclusions cannot be made concern- ing the effectiveness of individual treatment types. Sur- gical management of airway stenosis should be tailored to the type of stenosis, both in its severity and etiology.

Laryngoscope 127: January 2017

Lewis et al.: Surgical Treatment of Adult LTS

279