|Home | About | Journals | Submit | Contact Us | Français|
Objectives: Analysis of outcomes of a cohort of patients with esthesioneuroblastoma. Design: Retrospective cohort analysis. Setting: Patients presenting with esthesioneuroblastoma from 1994 to 2006 in a tertiary care academic medical center. Patients: Fifteen consecutive patients diagnosed as having esthesioneuroblastoma were treated during this time period using a subcranial resection. The mean follow-up is 75 months (range, 2 to 240 mos). Results: The overall survival was 100% and the overall disease-free survival was 49% and 24% at 5 and 15 years, respectively. Patients treated with radiation therapy following surgical resection had a 5- and 15-year disease-free survival of 83.3% compared with a 5- and 15-year disease-free survival of 26.7% and 0%, respectively, for patients whose initial treatment was surgery alone. The mean time to recurrence was 82.1 months. None of the patients had a decrease in Karnofsky Performance Score following subcranial resection. Conclusions: Patients with esthesioneuroblastoma whose initial treatment consists of surgical resection followed by radiation therapy have a longer disease-free survival than patients treated with surgery alone. However, initial treatment modality did not have an effect on survival. Long-term, close follow-up is necessary to identify recurrences, which can be treated with a high degree of success.
Esthesioneuroblastoma or olfactory neuroblastoma is a rare nasal cavity tumor first described over 80 years ago by the French physician Berger and colleagues.1 Representing only 3 to 6% of all nasal cavity cancers, only ~1000 cases have been described in the literature to date.2 Further complicating interpretation of the literature concerning esthesioneuroblastoma, Hyams grade IV esthesioneuroblastoma was often histologically confused with neuroendocrine carcinoma and other high-grade nasal cavity neoplasms, a problem that continued until relatively recent developments in histopathology enabled accurate diagnosis.3,4 Determination of optimal treatment protocols has also been limited by inconsistent treatment regimens between and within different institutions. Surgery, radiation, and chemotherapy have all been used in different combinations, with most institutions adopting a combined modality approach.5
The most common surgical approach is via craniofacial resection, an approach with proven efficacy and minimal side effects. The typical approach involves a lateral rhinotomy and bifrontal craniotomy with possible Weber-Ferguson facial flaps. The neurosurgical team approaches the lesion from above while the head and neck surgical team approaches the lesion from below. Despite the success of this technique, the poor cosmesis from the facial incisions and the functional status changes that can result from retraction of the frontal lobes can result in significant morbidity.6 An alternative to craniofacial resection is the transglabellar-subcranial approach. Initially described by Raveh and associates in 1988,7 the technique provides a more direct approach to the anterior skull base by disarticulating the nasal root with the bifrontal craniotomy through a bicoronal incision.8,9,10,11,12,13 In addition to avoiding frontal lobe retraction, this approach allows the malignancy to more easily be removed en bloc and spares the patient any facial incisions.
This article reviews our series of patients with esthesioneuroblastoma and our experience using the transglabellar-subcranial approach in the treatment of these patients.
The records of all patients with esthesioneuroblastoma at the University of Michigan were retrospectively reviewed. The patient population was obtained from the skull base registry maintained by the Department of Otolaryngology–Head and Neck Surgery, and University of Michigan Institutional Review Board approval was obtained before the study was initiated. All patients were staged using clinical and radiographic findings according to the Kadish classification system and treated surgically with a combined transglabellar-subcranial approach by a surgical team consisting of members from the Otolaryngology-Head and Neck Surgery and Neurosurgery departments. Statistical analysis was performed using the Kaplan-Meier technique. The follow-up time was defined as the time from surgery to recurrence of disease or the last clinical evaluation.
The transglabellar-subcranial approach has been previously described.14,15 A bicoronal incision is made and the flap carefully elevated while preserving the pericranium, which is later used as a lining for the skull base defect. After elevation of the flap is complete, bone cuts are made in the frontal bone and nasal root. The amount of fontal bone, orbital rim, and nasal root that are removed is determined by the amount of exposure that is necessary. The distal 3 to 5 mm of distal nasal bone are safeguarded to preserve the internal nasal valves. Prior to removal of the bone flap, rigid fixation miniplates are shaped and positioned to facilitate later bone flap repositioning. Prior to completion of the frontal bone cuts, the flap is fractured from the crista galli and the bony nasal septum. The bony cuts are then finished and the flap is removed, revealing a direct view into the nose, ethmoids, anterior fossa dura, and, if necessary, the posterior sphenoid, pituitary fossa, palate, and nasopharynx.
After exposure is obtained, the lesion is resected. If possible, the olfactory elements are preserved. In most cases, en bloc resection can be accomplished while preserving at least unilateral olfaction. After resection, the anterior fossa is sealed from the sinonasal cavity by placement of the previously raised pericranial flap. Often, the medial canthal ligaments are disrupted during the removal of the bone flap. These can be sutured to the opposite frontal bone using the centripetal suspension technique.14 The frontal sinuses can be stented open15 or obliterated.14
Fifteen patients were identified with histologically confirmed esthesioneuroblastoma who were treated between 1994 and 2007. The patients ranged in age from 19 to 71 years (Fig. 1) with 4 women and 11 men. Thirteen of the patients were Caucasian with one patient of Asian ancestry and one of Hispanic ancestry. Most patients presented with nasal obstruction (eight patients, 53%) or epistaxis (six patients, 40%). Other symptoms included asymptomatic nasal mass noted on exam in four (27%) patients, and one patient (7%) with each of the following presenting complaints: cushingoid appearance, hyponatremia, sinusitis, proptosis, ecchymosis, epiphora, hyposmia, and behavioral changes due to frontal lobe involvement (Fig. 2). At the time of presentation, one patient had disease in the neck. No patient presented with a tumor isolated to the nasal cavity (Kadish stage A). Thirteen patients presented with disease involving at least one paranasal sinus (Kadish stage B) and two patients with disease that extended beyond the paranasal sinuses (Kadish stage C; Fig. Fig.33).
The initial treatment was surgery alone for nine patients, surgery followed by postoperative radiation treatment in five patients, and surgery followed by postoperative chemotherapy and radiation in one patient (Fig. 4). All patients underwent complete gross surgical and radiographic excision of tumor. Two patients had positive microscopic surgical margins. The mean and median follow-up times were 75 and 64 months, respectively, with a range of 2 to 240 months. All patients are still living. Eleven patients are currently alive without disease, one patient is alive with disease, and three patients are alive with an unknown disease status.
Eight patients had recurrence of disease (Fig. 5). Seven of these eight patients recurred locally and three had regional recurrence in the neck. One of the two patients with positive surgical margins recurred. Of the three patients with regional recurrence, two also had local recurrence. Interestingly, the patient with isolated regional recurrence was the only one of the eight patients who recurred that received immediate postoperative radiation therapy (the neck was not a target of the radiation treatment). Of the eight patients who recurred, all but one received surgery alone as the initial treatment. All recurrences were treated with surgical resection followed by radiation therapy with the exception of one patient with isolated regional disease who was treated with neck dissection alone.
The mean time to recurrence was 82.1 months (95% confidence interval, 48.8 to 115.4 mos) and the overall recurrence-free survival was 49% at 5 years and 24% at 15 years (Fig. 6). There was no difference in disease recurrence between genders. Patients who presented with Kadish stage B had 5- and 15-year disease-free survival of 57%, whereas Kadish stage C patients had 5- and 15-year disease-free survival of 30% and 0%, respectively (Fig. 7).
Patients treated with surgery and postoperative radiation therapy fared better than those treated with surgery alone (Fig. 8). Patients who were treated with surgery alone had a mean recurrence time of 57.5 months (95% confidence interval, 17.0 to 98.1 mos), a 26.7% 5-year recurrence-free survival, and a 0% 15-year recurrence-free survival. Patients whose initial treatment was surgical resection followed by radiation therapy had a mean recurrence time of 122 months (95% confidence interval, 79.1 to 164.9 mos) and 5- and 15-year recurrence-free survival of 83.3%. Two patients received chemotherapy and each received this treatment at outside hospitals. One of the patients received three cycles of cisplatin and etoposide postoperatively in combination with radiation therapy. The second patient had esthesioneuroblastoma in combination with neuroendocrine carcinoma and received chemotherapy and radiation therapy postoperatively.
Functionally, patients who underwent subcranial resection maintained their preoperative functional status at 1, 6, and 12 months postoperatively (Table 1). One of the patients who recurred received whole-brain radiation and had a subsequent decline in Karnofsky Performance Score.
A total of nine complications were encountered. Two patients had cerebrospinal fluid leaks that resolved with conservative treatment, two patients had transient epiphora, two patients had postoperative diplopia, and one patient each had hydrocephalus, non-tension pneumocephalus, and a wound infection. The patient with hydrocephalus had a complicated medical history with a prolonged postoperative course and eventually required ventriculoperitoneal shunt placement for treatment of the hydrocephalus. The non-tension pneumocephalus was induced by nose-blowing and developed after the patient was discharged from the hospital. This patient was successfully treated with aspiration of the air and application of a compressive dressing. There was one patient who developed postoperative tension pneumocephalus and was treated successfully with tracheostomy and aspiration of the air.
The University of Michigan experience with esthesioneuroblastoma was previously documented.16 The current study documents our use of the subcranial approach for excision of these tumors. All patients in the study had histologically confirmed esthesioneuroblastoma. All other high-grade nasal cavity tumors, which have a high degree of microscopic similarity to esthesioneuroblastoma,3,4 were excluded.
The demographic profile of our patient population showed a unimodal age distribution with a peak incidence in the fifth decade. There were approximately three times more men than women. The finding of a unimodal age distribution and a higher incidence in men stands in contrast to traditional teachings on esthesioneuroblastoma, which report a bimodal age distribution and equal male and female incidence.17,18,19,20 Other reports have recently documented a similar age and gender distribution pattern.20,21,22,23
The overall survival in our series at both 5 and 15 years' follow-up is 100%. A recent review by Jethanamest et al22 identified a cohort of 311 patients diagnosed with esthesioneuroblastoma from the Surveillance, Epidemiology, and End Results tumor registry. In this group of patients, the overall survival was 62% at 5 years and 45% at 10 years. A meta-analysis by Dulguerov and associates24 demonstrated a 45% overall 5-year survival rate in the 26 studies that were included in the review from 1990 to 2000. Other recent studies have demonstrated an overall 5-year survival of 60 to 100%.21,25,26
A majority (8/15) of our patients experienced recurrence and the overall 5- and 15-year recurrence-free survival was 49% and 24%, respectively. Comparatively, the 5-year disease-free survival rates found in the recent literature are in the range of 33 to 87%5,21,25,26,27 This report also supports the conclusion of previous studies20,21,22,26 that patients with less-invasive disease fare better than those with more advanced disease (Fig. 7). In addition, we found that patients who received postoperative radiation therapy had a much improved 5- and 15-year disease-free survival (83.3% at both time points) compared with those initially treated with surgery alone (5- and 15-year disease-free survival of 26.7% and 0%, respectively). This is consistent with other studies that have shown improved outcomes with multimodality therapy.20,21,22,24,25,26,27 The finding of a poorer disease control rate without postoperative radiation is especially meaningful given that the surgical approach used in this series, the subcranial approach, allows en bloc resection of tumor, which results in at least a theoretical increase in the probability of obtaining negative surgical margins. We feel that this adds to the growing body of evidence that low-grade disease or negative surgical margins are not sufficient criteria to avoid the recommendation of postoperative radiation therapy.5,18,19,21,22,27,28,29,30
At our institution, chemotherapy is not routinely recommended as part of an initial treatment regimen for esthesioneuroblastoma. In this series, one patient, who had neuroendocrine carcinoma in combination with esthesioneuroblastoma, received chemotherapy postoperatively and another patient received postoperative chemoradiation therapy due to recurrent local disease. Both of these patients are currently alive without disease. The debate of the utility of chemotherapy for treatment of esthesioneuroblastoma continues in the literature. There are many reports of at least partial responses to chemotherapy.5,20,27,31,32 However, most studies on the effectiveness of chemotherapy have been conducted preoperatively or for salvage purposes. Many of these patients also received concurrent radiation therapy, making it difficult to determine whether the response is due to chemotherapy, radiation, or both. Despite these concerns, the experience at the University of Virginia, where preoperative radiation therapy and chemotherapy using cyclophosphamide, vincristine, and, in selected patients, doxorubicin, is part of the multimodality protocol for treating esthesioneuroblastoma, is strong evidence that this approach leads to improved disease-free survival. Their 5- and 15-year disease-free survival rates are the highest in the literature at 86.5% and 82.6%, respectively.5
Due to the high recurrence rate with esthesioneuroblastoma, salvage therapy is an important consideration. Our standard approach to recurrence is surgical excision followed by radiation therapy. Most of our recurrences were local in nature and three patients had recurrence of disease that presented in the neck. Two of these patients were treated with neck dissection and radiation and one of the two is currently disease-free. One patient was treated with neck dissection alone and is also currently disease-free. There were not any patients with distant metastatic disease. Our results with salvage therapy demonstrate the highly successful nature of treating recurrences of esthesioneuroblastoma and the need for long-term follow-up of patients with this disease.17,21,33 In addition to the usual postoperative clinical examinations and annual chest X-rays, our usual post-treatment follow-up routine is MRI scan at 4-month intervals for the first 3 years, then at 6-month intervals for the next 3 years, and then every 9 to 12 months thereafter.
The complications we experienced are similar to others described in the literature.5,18,19,20,21,24,25 However, we believe that the subcranial approach offers several advantages over traditional craniofacial approaches, including a more direct approach to the tumor allowing for en bloc tumor removal, improved cosmesis associated with avoidance of facial incisions, and the avoidance of frontal lobe retraction and its resultant detrimental effects on the patient's postoperative functional status. The ability of patients to maintain preoperative functional status is an important quality of life consideration as recently demonstrated by Woertgen et al.6 In this study by Woertgen, only 45% of patients were able to resume their previous occupation following surgical excision using combined transfacial and transcranial approaches. Our finding that Karnofsky Scores after surgical treatment were unchanged from preoperative scores supports the use of the subcranial approach over transfacial and transcranial approaches to maintain postoperative function status.
This article describes our experience treating esthesioneuroblastoma at the University of Michigan. The overall survival of our 15 patients is 100%. Eleven patients are alive without evidence of disease, three patients are alive with an unknown disease status, and one patient is alive with disease. The 5- and 15-year recurrence-free survival is 49% and 24%, respectively. Surgery followed by postoperative radiation therapy provided patients with the highest overall rate of disease-free survival. The role of chemotherapy in the treatment of these tumors is still evolving, but is most likely best applied to high-grade disease, recurrent disease, or unresectable disease. Close postoperative monitoring is important in order to detect recurrences and to offer treatment and salvage therapy. Salvage therapy for recurrences consists of surgical excision followed by postoperative radiation therapy, which offers a high rate of success. Further study is needed to establish and clarify treatment protocols.