Abstract
Purpose
The purpose of the study was to review a single-institution experience with endoscopic resection of sinonasal undifferentiated carcinoma (SNUC).
Materials and Methods
Thirteen patients underwent treatment of SNUC between January 2002 and July 2009. Retrospective data were collected including demographics, tumor characteristics, surgical strategy, adjuvant therapies, local and regional recurrence, distant metastasis, overall survival, and disease-free survival.
Results
The mean age was 51.8 years. The most common tumor stage at presentation was T4 (92%). Seven patients (53%) were treated with minimally invasive endoscopic resection (MIER) with negative intraoperative margins. Endoscopic anterior skull base resection was performed in 5 patients, and endoscopic-assisted bifrontal craniotomy was performed in 1 patient to clear the superior tumor margin. Six patients received pre- or postoperative chemoradiation. One patient underwent palliative chemoradiation, and one patient underwent open craniofacial resection. In the MIER group, simultaneous local and regional recurrence was observed in 1 patient (14%) after 30 months. Distant metastases were observed in 2 other patients (28%) without local or regional recurrence. All 3 patients with recurrences died of their disease. The remaining 4 patients were clinically, endoscopically, and radiographically free of disease, resulting in overall and disease-free survival rates of 57% with mean follow-up of 32.3 months.
Conclusions
These preliminary data suggest a potential role for MIER in the comprehensive management algorithm of SNUC in appropriately selected patients. Patient outcomes including local and regional recurrence, distant metastases, and overall and disease-free survival were comparable to a treatment strategy using traditional craniofacial resection.
Level of evidence
2b.
1
Introduction
Sinonasal undifferentiated carcinoma (SNUC) is a rare and aggressive neoplasm of the head and neck first described as a distinct clinical entity by Frierson et al in 1986. Arising in the paranasal sinuses, this malignancy occurs near critical structures and usually presents with locally advanced disease and carries high rates of regional and distant metastases. Because of the rarity of SNUC and paucity of controlled clinical studies, no consensus exists regarding optimal treatment . Multimodality therapy, including chemotherapy, radiotherapy, and surgery, has been widely used in an attempt to improve survival and disease control . Two-year survival rates among patients treated in this fashion in reported series to date range from 25% to 67% .
Traditionally, malignancies involving the paranasal sinuses and the adjacent anterior skull base (ASB) have been treated with craniofacial resection (CFR). This approach offers wide exposure, but is associated with significant morbidity and even mortality . Furthermore, CFR involves external incisions, with potential to alter physical appearance and affect patient quality of life . Endoscopic approaches facilitate diagnosis and evaluation of pathologies affecting the ASB and paranasal sinuses. Increasingly, endoscopic techniques are being adapted for resection of both benign and malignant tumors involving the paranasal sinuses and ASB . With accrued experience, indications for endoscopic resection of skull base neoplasia have been expanded with improved outcomes . The objective of this study is to evaluate the role of minimally invasive endoscopic resection (MIER) in the multimodality treatment paradigm of SNUC.
2
Materials and methods
Retrospective chart review was performed on all patients undergoing treatment of SNUC between January 2002 and July 2009 at the Cleveland Clinic Head and Neck Institute. The Cleveland Clinic’s Institutional Review Board approved this study. Data collected included patient demographics, tumor characteristics (primary site, TNM staging), treatment modalities including surgical management (intraoperative findings, surgical margins, complications), as well as the use of adjuvant and neoadjuvant therapy. Individualized treatment regimens were formulated based upon recommendations from the institutional head and neck multidisciplinary tumor board. Patient outcomes were assessed, including recurrence rates, survival (overall and disease-free), and long-term status (disease-free, alive with disease, dead of disease, dead from other causes).
Thirteen patients with SNUC were managed in the study period. Two patients were excluded from survival analyses because they did not fall into 1 of the 2 treatment groups studied (MIER/chemoradiation or definitive chemoradiation only). One of these was treated with palliative radiation, and the other patient underwent CFR and postoperative chemoradiation.
2.1
Statistical analysis
Kaplan-Meier curves were used to calculate overall and disease-free survival at 2 years for all patients and were also stratified based upon treatment type. For all estimates, 95% confidence intervals were calculated. Analyses were performed using R software (version 2.8; Vienna, Austria).
2
Materials and methods
Retrospective chart review was performed on all patients undergoing treatment of SNUC between January 2002 and July 2009 at the Cleveland Clinic Head and Neck Institute. The Cleveland Clinic’s Institutional Review Board approved this study. Data collected included patient demographics, tumor characteristics (primary site, TNM staging), treatment modalities including surgical management (intraoperative findings, surgical margins, complications), as well as the use of adjuvant and neoadjuvant therapy. Individualized treatment regimens were formulated based upon recommendations from the institutional head and neck multidisciplinary tumor board. Patient outcomes were assessed, including recurrence rates, survival (overall and disease-free), and long-term status (disease-free, alive with disease, dead of disease, dead from other causes).
Thirteen patients with SNUC were managed in the study period. Two patients were excluded from survival analyses because they did not fall into 1 of the 2 treatment groups studied (MIER/chemoradiation or definitive chemoradiation only). One of these was treated with palliative radiation, and the other patient underwent CFR and postoperative chemoradiation.
2.1
Statistical analysis
Kaplan-Meier curves were used to calculate overall and disease-free survival at 2 years for all patients and were also stratified based upon treatment type. For all estimates, 95% confidence intervals were calculated. Analyses were performed using R software (version 2.8; Vienna, Austria).
3
Results
3.1
Patient demographics
The mean age for the 13 patients was 51.8 years (range, 16–78). The male to female ratio was 1.2:1. The mean follow-up time was 23.3 months (range, 3–62 months).
3.2
Tumor characteristics
Diagnosis was based on histopathologic review by dedicated head and neck pathologists at the Cleveland Clinic. No specimens had evidence of neuroendocrine, squamous, or glandular differentiation. Table 1 reviews tumor site, stage, therapy, and survival status. Using the American Joint Committee on Cancer criteria for tumor staging, 12 of 13 patients had T4 lesions, whereas 1 patient was staged as T1 tumor limited to the maxillary sinus. Six tumors (46%) were primarily left sided, 3 (23%) were primarily right sided, and the remaining 4 (31%) were bilateral at presentation.
| Pt | Age | Tumor site | TNM stage | Surgical therapies | Adjuvant therapies | Recurrence | Additional procedures | Follow-up (mo) | Status at last follow-up |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 77 | Left orbitoethmoid with cribriform plate and dural involvement | T4bN0M0 | None | Palliative XRT | None | None | 5 | DOD |
| 2 | 39 | Left orbitoethmoid with infratemporal fossa and pterygopalatine fossa involvement | T4N0M0 | MIER | Preop cisplatin and etoposide, postop XRT | Local recurrence | None | 34 | DOD |
| 3 | 34 | Left orbitoethmoid with dural and brain involvement | T4bN0M0 | MIER with bifrontal craniotomy | Preop cisplatin and etoposide, preop XRT | Distant recurrence | None | 7 | DOD |
| 4 | 39 | Left frontal, orbitoethmoid with cribriform plate and dural involvement | T4bN0M0 | None | Definitive cisplatin and etoposide, XRT | Regional recurrence | None | 33 | AWOD |
| 5 | 73 | Right sphenoethmoid with lamina papyracea and cribriform plate involvement | T4bN0M0 | MIER | Preop cisplatin and etoposide, XRT | None | None | 62 | AWOD |
| 6 | 78 | Right maxillary, orbitoethmoid, with orbital, dural, infratemporal fossa involvement | T4bN0M0 | CFR with orbital exenteration | Postop cisplatin and etoposide, XRT | None | None | 5 | AWD |
| 7 | 49 | Left sphenoid, frontal, and orbitoethmoid with dural and cribriform plate involvement | T4bN0M0 | MIER | Postop cisplatin and etoposide, XRT | None | None | 24 | AWOD |
| 8 | 16 | Bilateral ethmoid with dural and brain involvement | T4bN0M0 | MIER | Preop cisplatin and etoposide, XRT | None | None | 17 | AWOD |
| 9 | 51 | Bilateral sphenoid, ethmoid | T4N0M0 | MIER | Postop cisplatin and etoposide, XRT | Distant recurrence | None | 28 | DOD |
| 10 | 51 | Bilateral frontal, ethmoid with dural and brain involvement | T4bN0M1 | None | Definitive cisplatin and etoposide, XRT. | None | None | 8 | DOD |
| 11 | 76 | Left maxillary | T1N0M0 | MIER | Postop XRT | None | None | 54 | AWOD |
| 12 | 47 | Bilateral sphenoid with cribriform, dural and brain involvement | T4bN2cM1 | None | Definitive cisplatin and etoposide, XRT | None | None | 21 | AWD |
| 13 | 44 | Right maxillary, ethmoid | T4N0M0 | None | Definitive cisplatin and etoposide, XRT | None | None | 9 | AWOD |
The most common sinuses affected in descending order were the ethmoid in 11 of 13 tumors (85%), sphenoid in 8 (62%), frontal in 6 (46%), and maxillary in 5 (38%). Eight tumors (62%) involved the cribriform plate, with dural involvement in 5 cases (38%) and brain involvement in 3 (23%) of these cases. Seven patients (54%) had involvement of the lamina papyracea, with 4 (31%) of these patients having periorbital involvement and 2 (15%) with orbital involvement at presentation. Three tumors (23%) involved both the pterygomaxillary fossa and infratemporal fossa. Two patients (15%) had involvement of the clivus, and 1 patient (8%) had nasopharyngeal extension.
At presentation, 2 patients had distant disease in the form of liver metastases; and 1 of these 2 patients also had regional disease with bilateral cervical adenopathy.
3.3
Surgical management
Eight patients (62%) underwent surgical treatment. Seven patients underwent MIER for definitive tumor management. Image guidance was used in all 7 cases. One patient in this group required bifrontal craniotomy to clear the superior tumor margin because of intracranial extension. One additional patient underwent open CFR consisting of extended maxillectomy, open sphenoidotomy, and ethmoidectomy with orbital exenteration.
Six (86%) of 7 patients treated with MIER underwent formal endoscopic ASB resection. Reconstruction was performed based on the preference of the attending surgeon in a multilayered fashion in all cases . Three patients were noted to have cerebrospinal fluid leaks intraoperatively, with 2 patients having subarachnoid lumbar drain placement for 3 to 5 days. Neurosurgical assistance was required during 2 cases, including patient 3 requiring bifrontal craniotomy. Orbital decompression was necessary in 4 patients, and optic nerve decompression with subtotal resection of the clivus was performed in 1 patient. There were no perioperative complications or delayed cerebrospinal fluid leaks.
3.4
Chemotherapy and radiation therapy
All patients treated in this study received chemotherapy (CTX) and/or radiation therapy (XRT). The most common chemotherapeutic regimen consisted of cisplatin (60 mg/m 2 given once) usually concurrently with etoposide (120 mg/m 2 for 3 days repeated every 3 weeks for 2–5 total cycles).
All patients received radiotherapy with curative intent to the primary site, with the exception of patient 1 who underwent palliative XRT only. Daily fractions (200 cGy) were administered to a total dose of 56 to 70 Gy (mean, 59.4 Gy). Three patients (27%) underwent elective radiation to the neck from 46 to 60 Gy (mean, 53.3 Gy), with 1 receiving preoperative and 2 postoperative therapy. The patient undergoing CFR received postoperative intensity-modulated radiation therapy to the primary site.
Four patients underwent concurrent CTX/XRT as definitive treatment without surgical intervention. Two of these patients received planned preoperative CTX/XRT; however, because no evidence of disease was present on posttreatment imaging or diagnostic endoscopy, surgery was deferred. One of these 2 patients was found to have regional recurrence at 10 months, which was treated with radiation and modified radical neck dissection and was alive without evidence of disease at 23 months after recurrence. The other patient is currently under surveillance without evidence of disease at 9 months. Another patient upon completion of planned preoperative CTX/XRT was found to have unresectable local and distant metastases and died of his disease. The last of the 4 patients was started on CTX and XRT because of large, unresectable intracranial disease. Distant and regional metastases were discovered during treatment, and the patient soon thereafter died of disease.
Six of 7 MIER patients had concurrent CTX/XRT, with 3 (43%) of 7 patients receiving preoperative therapy and 3 (43%) receiving postoperative therapy. One of 7 (patient 2) underwent preoperative CTX followed by postoperative XRT. The sole patient with a T1N0M0 tumor (patient 11) underwent only postoperative radiation without CTX because of advanced age, comorbidities, and limited extent of disease.
The only reported complications of chemotherapeutic or radiation treatments were transient myelosuppression in 2 patients.
3.5
Survival data
Survival and recurrence data were compiled for the 11 patients undergoing either definitive CTX/XRT or MIER. Four patients (36%) recurred following definitive treatment. For the entire cohort, recurrences were found to occur locally in 1 case (9%), regionally in 2 cases (18%), and distantly in 2 cases (18%).
Of the MIER group, simultaneous local (left parapharyngeal space) and regional (left level II lymph node) recurrence was observed in 1 patient (14%) after 30 months. This was treated with adjuvant radiation to the primary site and neck following unilateral selective neck dissection. Distant metastases were observed in 2 other patients (28%) without local or regional recurrence. These were observed as bilateral pulmonary metastases in 1 patient (patient 3) at 6 months and right iliac bone metastasis in the other (patient 9) at 10 months. All 3 patients in the MIER group with recurrences died of their disease at an average of 23 months following therapy.
At the time of last follow-up, 6 (54%) of 11 patients were alive without evidence of disease, 3 (27%) were dead from disease, and 1 patient (9%) was alive with disease. Within the subgroups ( Table 2 ), 3 patients treated with MIER had died of disease; and the remaining 4 patients were clinically, endoscopically, and radiographically free of disease, resulting in overall and disease-free survival rates of 57% at mean follow-up of 32.3 months.
