European Archives of Oto-Rhino-Laryngology

, Volume 264, Issue 12, pp 1419–1424

Surgical management of sinonasal inverted papillomas through endoscopic approach


    • Department of Otorhinolaryngology and head and neck surgery, Huriez HospitalUniversity of Lille
    • Service d’ORL et de chirurgie cervico-faciale, Pr. Chevalier
  • E. Arzul
    • Department of Otorhinolaryngology and head and neck surgery, Huriez HospitalUniversity of Lille
  • J. A. Darras
    • Department of Otorhinolaryngology and head and neck surgery, Huriez HospitalUniversity of Lille
  • D. Chevalier
    • Department of Otorhinolaryngology and head and neck surgery, Huriez HospitalUniversity of Lille

DOI: 10.1007/s00405-007-0401-2

Cite this article as:
Mortuaire, G., Arzul, E., Darras, J.A. et al. Eur Arch Otorhinolaryngol (2007) 264: 1419. doi:10.1007/s00405-007-0401-2


We defined a standardized approach to surgery of sinonasal inverted papillomas (IP) for adequate and safe resection. A cohort of 65 patients treated from January 1995 to December 2005 at a single institution was retrospectively analyzed (mean follow-up: 28 months; range 1–132). The extension of the tumor was evaluated on clinical findings and computed tomography (CT) scan and/or resonance magnetic imaging (RMI). External and endoscopic surgical approaches were compared according to tumor extension, rate of local recurrence. Univariate analysis was used to review the impact on disease-free survival of factors related to the histopathological findings and the treatment. Endoscopic (alone or combined with transantral approach) and external surgery were used in 46 patients (71%) and 19 patients, respectively. Endoscopic approach (34/46) was performed to control IP in the nasal fossa, the ostiomeatal complex, the sphenoid sinus. It was combined with Caldwell-Luc procedure (12/46) for tumor extent into the lateral part of the maxillary sinus. The mean time for recurrences to occur was 19 months with range of 5–35 months. The rate of local recurrence was 17.6% (6/34) in endoscopic approach alone, 8.3% in endoscopic approach combined with a Caldwell-Luc procedure and 15.8% (3/19) in external approach. Tumor extension, excision with safe margins, associated malignancy or dysplasia have no significant impact on disease-free survival regardless of surgical procedure. On the basis of imaging evaluation and peroperative view of tumor extent, we propose a surgical strategy in which endoscopic approach could be used on the first attempt by trained surgeons. RMI is very useful to determine acute extent of the disease.


Inverted papillomaEndoscopic surgeryResonance magnetic imagingRecurrence


Inverted papilloma of the sinonasal tract is a rare and benign tumor. It arises from ectodermally derived nasal and paranasal sinus mucosa known as the Schneiderian membrane in the lateral nasal wall. It accounts for 0.5–4% of the nasal tumors [3, 13]. In 1991, the World Health Organization published a classification of the nasal papillomas which are divided into three histopathological types: inverted papilloma, exophytic papilloma and columnar cell papilloma [13]. IP is more frequently encountered in 70% of the sinonasal papillomas [3, 15]. Etiopathogenicity is still hypothetic. Allergy, chronic rhinosinusitis, pollution, tobacco smoking could play a role in IP arising without strong evidence [1]. A possible viral origin by Human Papilloma Virus has been repeatedly reported by several authors [6].

IP management remains a matter of controversy. The treatment is by surgical resection. In spite of benign histological features, IP is known to be aggressive with local bone erosion and paranasal extent. It is at high risk of local recurrence and associated malignancy is found in about 9.1% of cases [8]. Regarding this aggressive nature, most authors to date have recommended a wide total resection. External approach through lateral rhinotomy is supposed to be the gold standard of treatment. With the recent improvement of endoscopic techniques, endonasal resection is introduced by other authors as an alternative surgical procedure with less morbidity than external approach.

By evaluating our results on recurrence rate, the aim of this study is to establish a classification based on the location and the anatomical extension of IP. This classification could help to determine properly indications for each surgical technique especially by pointing out indications of endoscopic procedure.

Materials and methods


From January 1995 to December 2005, 65 patients with inverted papillomas of the sinonasal tract were treated by surgery in the Department of Otolaryngology in Lille. There were 51 males and 14 females with a mean age of 57 years (range 18–87 years). All patients had biopsy-proven inverted papilloma. The patients with associated malignancy at diagnosis were excluded from the study. Data from these patients were not analyzed.

Tumor extension

The location and the extension of IP were evaluated in the preoperative course on clinical examination with nasal endoscopes and imaging. CT scan was performed in 56 patients (86%) and RMI in 15 patients (23%). CT scan and RMI were both performed in seven cases. Before surgery, examination with rigid nasal endoscopy was done under general anesthesia. The root of the tumor was determined in each case.


The surgical approach was chosen on the basis of tumor location and skills of the surgeon. Endoscopic procedure was performed in 46 cases (71%). It consisted of adequate ethmoidectomy alone in 30 patients, with endoscopic medial maxillectomy in four patients, with combined approach through Caldwell-Luc transantral procedure in 12 patients. Neither modified Lothrop procedure nor combined approach with frontal osteoplastic flap was used in our study. External approach was performed in 19 cases (29%): lateral rhinotomy in five patients, lateral rhinotomy with medial maxillectomy in nine patients, Caldwell-Luc in three patients and midfacial degloving procedure in two patients. Surgical specimens were sending to acute histopathological analysis (quality of margins, associated dysplasia or malignancy, associated polyps).

The patients were discharged at home with a mean of 2 days.


The patients were followed-up by endoscopic controls every 3–6 months. In the last 2 years of the study, RMI as reference imaging was performed at 3 months in the postoperative course. The median of follow-up was 28 months with range from 1 to 132 months. One patient left the country, 1 month after surgery.

Statistical methods

The univariate analysis of prognosis factors with impact on disease-free survival was performed by comparing actuarially predicted curves with the log rank test. A P value less than 0.05 was deemed statistically significant. Types of surgical approach, quality of histopathological margins, associated dysplasia or malignancy regardless of surgical procedure were evaluated. Data were analyzed with EPI6FR software version 6.0 and STATISTICA software version 4.0.



The most frequent presenting symptom was unilateral nasal obstruction in 53 cases (81%). Post-nasal drip was described in 17 cases, epistaxis in 6 cases. Ten patients (15%) were referred to our institute for a local recurrence with first treatment at an outside hospital.

Tumor extension

There were 34 left-side lesions and 31 right-side lesions. No case of bilateral IP was described. On the 56 preoperative CT scan performed, no specific sign was described. Bone erosions in ethmoid were found in 17 patients (26%), deformation of the ostiomeatal complex in 7 patients (10, 8%), and irregular sclerosis with internal hyperdensity in 3 cases (4.6%). On the 15 preoperative RMI performed, IP showed homogeneous enhancement with intravenous gadolinium. T2-weighted MRI was very accurate in distinguishing IP (intermediate signal) from adjacent inflammatory change (very high signal). In seven patients, both CT scan and MRI were performed. CT scan overestimated tumor extension in four cases. It did not distinguish IP from retained secretion in both frontal and maxillary sinus in two cases, in frontal sinus alone in one case and in sphenoid sinus in one case (Fig. 1).
Fig. 1

Imaging of inverted papilloma. Coronal (upper left part) and axial (lower left part) CT scan showing appearance of inverted papilloma. The extension of the tumor was overestimated in the frontal and sphenoid sinus (black arrows) when we compare with T2-weighted RMI performed on coronal (upper right part) and axial (lower right part) views. RMI distinguish IP from adjacent inflammatory mucosa with retained secretions

The most common site of involvement for IP was the lateral nasal wall (middle turbinate, ostiomeatal complex, inferior turbinate) in 51 cases. Of the patients, 33 had involvement of the maxillary sinus, 21 had involvement of the anterior ethmoid, 12 the posterior ethmoid, 3 the frontal sinus. Two patients presented with IP of the nasal septum. Two patients with extranasal extension were noticed: one case through cribiform plate in extradural spaces, one case in retromaxillary soft tissues. On definitive histopathological findings, this last patient was suffering from an associated squamous cell carcinoma. No case of sphenoid extension was noticed. The root of the tumor was identified on the lateral nasal wall in 45 cases (69%), on the walls of the maxillary sinus in 11 cases.


Endoscopic approach was used mainly for IP in the nasal fossa, in the ostiomeatal complex, in the posterior ethmoid and the frontoethmoid area. In eight patients, this approach was enlarged to endoscopic medial maxillectomy to control the lateral part of the maxillary sinus. Combined procedure with endoscopic and transantral approaches were performed for IP with extension in the lateral, anterior and posterior part of the maxillary sinus. External approach with lateral rhinotomy was performed for IP with ethmoidal or frontoethmoidal involvement. Midfacial degloving procedure was used in only two cases with a tumor of the nasal septum and a tumor in the posterior part of ethmoid next to the sphenoid recess (Table 1). Cribiform plate plasty with turbinate mucosa was performed in four cases (three through lateral rhinotomy, one through endoscopic view).
Table 1

Distribution of surgical approaches (n = 65) according to tumor extension


Surgical approaches


Caldwell-Luc ± endoscopic

Lateral rhinotomy

Midfacial degloving


Tumor Involvement

Confined to the nasal fossa





Involving ethmoid sinus or the medial portion of the maxillary sinus





Involving the frontal or the sphenoid sinuses or the non medial portion of the maxillary sinus






Extension beyond the nasal cavity or the paranasal sinuses










During postoperative follow-up, three epistaxis occurred. A case of suppurative cavity, transient diplopia and infraorbital hypoesthesia were described, respectively, after three external approaches. Persistent crusting was noticed in 12 cases.

The global local recurrence rate was 15% (10/65). Three of these ten patients with local recurrence were referred to our department for a first recurrence. A first recurrence was not associated with high risk of a second recurrence (P = 0.17). The median delay of recurrence was 19 months (range 5–35 months). With the Kaplan–Meier method, the 3-year disease-free survival was estimated in our study to be 72.8%. The local recurrence rate was 15.8% (3/19) with external approach, 17.6% (6/34) with endoscopic approach alone and 8.3% (1/12) with endoscopic approach combined with a transantral procedure. Among these ten recurrences, two patients were treated in second procedure through endoscopic surgery with medial maxillectomy, two patients through Caldwell-Luc transantral approach, two patients through transnasal polypectomy, one patient with a Denker procedure, one patient with a lateral rhinotomy and one patient with a combined endoscopic/transantral approach. The last one was submitted to accurate follow-up without additional surgery. No second recurrence was noticed so far.

Prognosis factors

No significant impact of surgical approach on disease-free survival was noticed between endoscopic procedures alone and external procedures with the Log-rank test (P = 0.20). A macroscopic adequate resection had no predictive value on good local control (P = 0.58).

Analysis of surgical margins on specimen was difficult in cases of piecemeal resection. Meanwhile, no statistical link between safe margins and local recurrence was found. Associated cancer with IP was found in seven cases (11%): squamous cell carcinoma in three cases, carcinoma in situ in two cases, adenocarcinoma in one case and malignant melanoma in one case. External radiotherapy was performed in three patients as additional treatment. A second look surgical procedure was not scheduled in any case. The median follow-up for these patients with associated cancer was 31 months (range 1–84 months). Two of them with associated squamous cell carcinoma died: one from metastatic diffusion, one from intercurrent disease.

Associated mild dysplasia with IP was notified in five cases (7.7%). Associated cancer (P = 0.29) and dysplasia (P = 0.57) did not influence the local control.


With low incidence and non-specific presenting complaints, the diagnosis of IP is not so easy to perform [4]. The surgeon should be aware in case of unilateral nasal mass in adult with a past history of chronic unilateral nasal obstruction. Some cases have been described in childhood on histopathological analysis after endonasal large polypectomy [3].

IP is thought to be locally aggressive with a high risk of local recurrence in case of inadequate surgery [1]. The management of IP requires biopsy-proven tumors on preoperative course to ensure the finest mapping before surgery [15]. CT scan and RMI are both useful. CT scan shows anatomical landmarks and bone erosion of adjacent structures such as the skull base or the orbit [1, 11]. T2-weighted RMI is necessary to assess tumor boundaries by distinguishing IP from adjacent inflammatory mucosa with retained secretions especially in paranasal sinuses [1]. On the basis of histological findings and CT scan/RMI evaluation, surgery can be planned properly.

Radical excision through external approach was considered as the treatment of choice of IP. This assertion was supported by the high local recurrence rate of IP in the first series of conservative endoscopic surgery [8]. In the last decade, some authors tried to specify indications of the endoscopic approach with the advent of endoscopic sinus surgical techniques. On selected cases, they obtained a good local control with a recurrence rate of 12% for Lawson and 17% for Waitz and Wigand on series of 35 and 41 patients, respectively [1, 9]. In our study, the results of endoscopic approach were comparable. Moreover, these authors emphasized a better outcome of endoscopic surgery with short hospitalization, less bleeding and less injury than external procedures (facial scar, nasal vestibule stenosis, lacrymal duct injury) [1, 13]. Endoscopic surgery allows preservation of bony structures and safe mucosa. This framework avoids gateway to extranasal spaces in case of local recurrence.

The main topic is not controversy between external surgery and endoscopic surgery but definition of appropriate approach for each IP tumor. It is quite impossible to compare local recurrence rate of these two procedures. They are used for tumors of different size and location. In published studies, the endoscopic procedure is used for IP in the nasal fossa, the septum, the ostiomeatal complex, the medial part of the sinus maxillary, the posterior ethmoid and the sphenoid sinus [7, 9]. A good visualization of the lateral, inferior and posterior parts of the maxillary sinus is difficult through endoscopic view. Kamel described an endoscopic medial maxillectomy with complete resection of the lateral nasal wall to achieve exposure [4, 5]. A combined approach with Caldwell-Luc transantral procedure was also described in literature [2, 7, 14]. It was used mainly in our study to control the maxillary sinus with a very low recurrence rate. Concerning the frontal sinus and the frontoethmoidal recess, endoscopic approach with Lothrop procedure [1, 10, 12] or combined approach with osteoplastic frontal flap are proposed [14]. For extranasal IP, external approaches are always performed [13, 15]. In case of local recurrence, the endoscopic approach can be used in the same way. Long-term follow-up is recommended. In our study, recurrences occurred up to 35 months.

Regardless of surgical techniques, no prognosis factor of local control was pointed out in our study. The main rule is to identify the root of the tumor to ensure complete excision with safe mucosa margins with en-bloc resection [5]. Some authors described drilling out of underlying bone structures without clear evidence of improving local control [10, 14].

Several classifications for IP were described to allow comparison in literature. In 2000, Krouse proposed a staging based on histological findings of associated malignancy and location of the tumor evaluated by clinical examination and CT scan [7]. Biopsy-proven IP with associated malignancy is required to avoid misinterpretation in the choice of aggressive surgery for small size tumor of the nasal fossa. To help in the surgical management of IP, we describe a classification based on tumor location evaluated by clinical preoperative endoscopy, CT scan and RMI. Each item refers to specific anatomical location of IP (with distinguishing frontal and maxillary extension accessible to combined approach). A first line treatment is proposed for each item with a possible switch according to preoperative findings and surgeon skills (Table 2).
Table 2

Surgical strategy for resection of inverted papilloma

Anatomical extension (endoscopic and CT scan/IRM)

First line treatment procedure

Second line treatment procedure

Nasal fossa


Midfacial degloving or lateral rhinotomy



Anterior ethmoid (bulla, infundibular cells)

Endoscopic with medial maxillectomy

Midfacial degloving or lateral rhinotomy

Posterior ethmoid


Ostiomeatal complex, medial and posterior walls of the maxillary sinus


Sphenoid sinus


Lateral, inferior and anterior walls of the maxillary sinus

Endoscopic with medial maxillectomy

Midfacial degloving


Combined approach: endoscopic with Caldwell-Luc transantral procedure


Anterior ethmoid (frontal recess cells)

Endoscopic with Lothrop procedure

External frontal approach

Frontal recess

Combined approach: endoscopic with frontal osteoplastic flap

Lateral rhinotomy

Frontal sinus


Extranasal (cribiform plate, anterior skull base, orbit, soft tissues)

Lateral rhinotomy


Craniofacial procedure



In the management of inverted papillomas, tumors must be carefully delimitated on CT scan combined with RMI in order to clearly enhance the treatment planning. Biopsies with clear histopathological results should be available preoperatively. Endoscopic and external approaches are both available to achieve radical excision of the tumors with similar rates of recurrence. Careful case selection for each technique is mandatory. Endoscopic approach can be used in many cases by skilled and trained surgeons. Anyway, patients must be told of a possible external approach that will be decided preoperatively according to tumoral extent findings.

Copyright information

© Springer-Verlag 2007