2018 Section 5 - Rhinology and Allergic Disorders

Eur Arch Otorhinolaryngol (2016) 273:4343–4350

cleft in a series of 44 patients, we described ADC location in this area in 90 % of our patients. An accurate attention must be paid to this area during the surgical procedure to ensure a complete removal of the tumour. Although uni- lateral olfactory cleft resection technique is proposed by some authors [ 4 , 17 ], we believe that a bilateral resection offers a better control of tumoral extension next to the midline. Moreover, it allows a better support for a potential dura plasty. In this perspective, we perform a Draf 3 frontal sinusotomy to improve the anterior oncologic control of our resection. This procedure is also helpful to prevent frontal blockage in the healing process and to improve nasoendoscopy visualization during the clinical follow-up. ESS allows a complete mucosal resection as long as the tumour is not extended beyond some critical boundaries. T3 with laterally and anteriorly involvement of the bony structure of the maxillary sinus, T4a with superiorly and anteriorly frontal sinus extension, T4a with skin extension, and T4b with orbital invasion still require an open approach, whereas large involvement of the anterior skull base and the overlying dura is manageable through ESS. Intracranial resections by ESS represent a major progress in the surgical treatment of sinonasal tumours and allow extending the indications of surgery to older patients. Meanwhile, a strict selection of patients is always mandatory. LR is still a useful option in cases of T4 lesions with major extrasinusal extension. As a consequence, a thorough preoperative assessment of ADC location is mandatory before multidisciplinary decision. Our study showed that CT scan and MRI can overstate or understate the tumour extensions. The involvements of the olfactory cleft and/or the cribriform plate which are epicenter of many ADC are often misin- terpreted in the first place. Previous studies showed that a clear delineation of anterior skull base and dural infiltra- tions was difficult to achieve even with MRI. In a com- parative study of imaging findings and histopathological examination of patients with benign and malignant tumours, Gomaa et al. demonstrated that false positivity was observed for CT scan in the ethmoid sinus analysis [ 23 ]. Our results about malignant tumours are consistent with this study, as imaging sensibility for cribriform plate involvement was 65.5 % in our study. In a retrospective analysis of 20 consecutive patients with diffuse dural enhancement detected by MRI, River et al. showed that pathological evidence of leptomeningeal invasion was found in only one case [ 24 ]. In our study, imaging PPV for meningeal infiltration was 80 %. More than the CT scan and MRI findings, it is the intraoperative assessment of tumour extensions and origin that leads the surgeon to adequately tailor the surgical resection. Frozen sections have been proposed as a reliable approach [ 25 ]. It should be considered for anterior skull base tumours when radical

an almost comparable study design but over a longer period (21 years) and including some patients without adjuvant RT [ 17 ]. Our rates of local failure for LR (21.7 %) and ESS (20 %) are comparable with the previous studies [ 9 , 11 ]. We observed, respectively, five and four local failures in LR and ESS groups with a very variable delay (mean: 31.6 months (from 8.5 to 56.5 months). The mean time to local reccurence in GETTEC multicenter study was 28 months [ 2 ]. In our study, one patient treated with LR showed early local recurrence, before the 1st year of fol- low-up, even though the tumour was only staged pT3. Before being referred to our institution, the patient was thought to have benign condition, such as inverted papil- loma and underwent large resection for histological diag- nosis. Postoperative changes, such as altered anatomy and edema, may have made accurate delineation of the extent of disease and definitive surgical resection more difficult. We postulate that it may be responsible for the bad short- term outcome in this case. It is worth noticing that seven of the nine local failures were also treated with subsequent ESS. In patients with the previous surgery and/or RT, we postulate that ESS can also be considered as a viable sal- vage treatment option. Taken together, these results underline the oncological validity of ESS in the surgical management of ADC. Meanwhile some issues need to be raised. Piecemeal resection performed in ESS has been criticized by some authors when this procedure is applied with oncological perspectives. Although the concept of ‘‘en bloc’’ resection is difficult to apply even to transfacial resection [ 17 ], piecemeal resection does not seem to impair the local control [ 18 ]. Like Nicolai et al. [ 9 ], we advocate a com- plete centripetal ethmoid sinus resection even for limited tumours. Recent studies advocate for a bilateral eth- moidectomy to prevent local recurrences controlaterally, in particular in the cases of ascertained occupational exposure [ 19 , 20 ]. Our attitude is strengthened by observations of preneoplastic and neoplastic foci in macroscopically uninvolved ethmoid areas in particular in patients exposed to carcinogenetic dusts [ 21 ]. In our study, a contralateral involvement of the olfactory cleft was histologically described in 4 cT2 and 1 cT3 patients. In a series of ten ITAC of the sinonasal tract, Kennedy et al. identified intestinal metaplasia of the respiratory mucosa surrounding the neoplasm. It was accompanied with a phenotypic switch to an intestinal phenotype [ 22 ]. These results sug- gest that a multifocal origin of ITAC is possible. Further histopathological studies are required to support this hypothesis. ESS offers excellent visualization of tumours origin and extension. It provides easy access to the olfac- tory cleft, which is considered as the starting point of the ethmoid ADC in woodworkers [ 4 ]. Like Van Gerven et al. [ 11 ] who observed 84 % of involvement of the olfactory

123

254