Kaplan-Meier analysis of survival. Analysis is adjusted for disease-free survival.
Local relapse-free and distance relapse-free survival.
Knegt PP, Ah-See KW, vd Velden L, Kerrebijn J. Adenocarcinoma of the Ethmoidal Sinus ComplexSurgical Debulking and Topical Fluorouracil May Be the Optimal Treatment. Arch Otolaryngol Head Neck Surg. 2001;127(2):141-146. doi:10.1001/archotol.127.2.141
To report our experience with the management of adenocarcinoma of the ethmoidal sinuses using a regimen of surgical debulking and topical chemotherapy, to report long-term survival, and to compare our results with recently published series of patients undergoing craniofacial resection.
Review of prospectively collected data.
Tertiary cancer center.
Seventy consecutive patients with ethmoidal adenocarcinoma referred to the Department of Head and Neck Surgery, University Hospital of Rotterdam, Rotterdam, the Netherlands, between January 1976 and December 1997. Sixty-two patients were eligible for primary treatment.
Surgical debulking via an extended anterior maxillary antrostomy followed by a combination of repeated topical chemotherapy (fluorouracil) and necrotomy. Additionally 8 patients (13%) required radiotherapy for local recurrence; 1 patient required surgery for regional lymph node metastases.
Main Outcome Measures
Survival measured by the Kaplan-Meier method. Clinical complications related to the therapy.
There were no perioperative deaths. Complications did occur, such as temporary periorbital swelling (25 patients [40%]) and temporary cerebrospinal fluid leakage (5 patients [8%]). One patient (1.6%) developed meningitis. Adjusted disease-free survival at 2, 5, and 10 years is 96%, 87%, and 74%, respectively.
Our 23-year experience with a combination of surgical debulking and repeated topical chemotherapy for patients with adenocarcinoma of the ethmoidal sinuses leads us to believe that it represents the current treatment of choice for these patients for long-term disease-free survival.
PARANASAL SINUS tumors are rare forms of head and neck malignant neoplasms accounting for only 3% of head and neck cancers. Within this group exist various histological forms including squamous cell carcinoma, adenocarcinoma, and undifferentiated carcinoma. Epidemiologically adenocarcinoma has been associated with workers exposed to hardwood dust, with this association being first reported in 1965 by MacBeth.1 With improvements in surgical and anesthetic techniques, more radical surgical approaches have been developed, with the current choice being a 2-pronged approach from above and below the tumor, namely, the craniofacial resection (CFR).
Craniofacial resection has been adopted worldwide as the standard therapy for tumors of the paranasal sinuses involving the anterior skull base.2- 4 Since its publication by Ketcham et al5 in 1963, the technique has been refined and consequently morbidity and mortality have been reduced.
For some types of histopathology like esthesioneuroblastoma, CFR offers a better prognosis than other forms of treatment.6 Unfortunately, for adenocarcinoma of the ethmoidal sinus, CFR has not rendered significantly improved survival rates.2- 4,7 Lateral rhinotomy followed by radiotherapy has recently been shown to produce comparable survival results8 while primary radiation therapy alone remains an alternative for patients unfit for surgery.9,10
As regards the clinical outcome and survival data for these patients, therefore, the question still remains as to whether CFR is the best approach to these tumors. The aim of this article is to present our 23-year experience in the management of the patients with adenocarcinoma of the ethmoidal sinus complex treated in our department using a technique modified from that originally described by Sato et al11 in 1970.
In Rotterdam since 1976 we have treated all patients with adenocarcinoma of the ethmoid in a standard fashion. This consists of surgical debulking followed by topical chemotherapy combined with frequent necrotomy.12 In the period January 1976 through December 1988, these patients also received low-dose radiotherapy (ie, 2 × 2 Gy preoperatively and 5 × 2 Gy postoperatively). However, in consultation with our radiotherapy colleagues it was believed that this low-dose radiotherapy was insufficient to have any influence on the results. Therefore, in 1988, this practice was stopped and patients thereafter received only surgical debulking and topical chemotherapy with regular necrotomy.
Preoperative assessment included computer tomographic scanning to assess the extent of the tumor with particular attention given to the orbit and anterior cranial floor. Gross invasion through the dura to brain tissue or through the periorbital lining is a contraindication to this approach and consideration is then given to possible CFR. All pathologic sections were reviewed by our own pathologist to confirm the diagnosis of adenocarcinoma. It can often be difficult to ascertain whether the tumor has originated from the ethmoidal sinus or the maxillary sinus. All of our patients, however, had biopsy-proved evidence of ethmoidal sinus disease, while tumor in the maxillary sinus may or may not also have been present.
The surgical approach to the paranasal sinuses is via the anterior wall of the maxillary sinus as in a Caldwell-Luc procedure. The anterior wall, however, is more extensively opened up to within 2 mm of the inferior oribital rim, while taking care to preserve the infraorbital nerve. The lateral nasal wall, including the frontal process, medial wall of the maxilla, and the turbinates are removed. All sinuses including the maxillary, ethmoids, and sphenoid are meticulously cleaned. In essence a complete medial maxillectomy plus sphenoethmoidectomy according to Denker13 and as described by us14 is done. As much tumor as possible is removed, taking care to try to preserve important structures such as the dura and periorbit. In our experience these 2 structures have proved to be robust at resisting tumor invasion.
At the end of the procedure, areas in which tumor may have remained are liberally covered with a 5% fluorouracil emulsion (Efudix; Hoffmann-LaRoche, Basel, Switzerland). The cavity is then packed with 10 × 10-cm gauze swabs impregnated with 3% tetracycline hydrochloride in a petroleum jelly ointment.
Postoperative treatment consists of twice-weekly changes of the tetracycline packs and meticulous removal of necrotic material, followed by further fluorouracil application. This is performed in the outpatient clinic. The first postoperative change is performed after the patient has received intramuscular morphine. Subsequent visits require only acetaminophen (in Europe, paracetamol) (1 g, orally) 1 hour prior to the appointment. The dead tissue slough is easily identified as a painless gray tissue layer that can be removed with brisk suction.
Endoscopic surgery and subsequent endoscopic treatment has not been been adopted in this series. In our experience the open surgical technique provides an unhampered view of the sinuses, especially of the orbital floor and the zygomatic and alveolar recesses. The postoperative treatment can be performed with minimal burden for the patient via the large anterior antrostoma. Routine inspection of the cavity after treatment can, however, be performed endoscopically.
A total of 8 dressing changes and meticulous necrotectomies are performed. Further packing of the cavity is continued, however, for a further 4 weeks during the healing phase to prevent closure of the anterior maxillary sinus wall defect, thus providing a permanent portal for treatment and, importantly, for regular inspection. From the first postoperative day patients eat a normal diet. The tetracycline-impregnated gauzes protect the cavity during the early stages, while regular nasal douching keeps the cavity clean thereafter. Patients are advised to regularly rinse both the oral and nasal cavities after meals with an isotonic (0.9%) saline solution. Nasal douching can easily be performed using a 10-mL syringe filled with warm saline solution. Our patients tolerate this method of aftercare very well.
Between 12 and 16 weeks after the start of therapy an endoscopic inspection of the cavity is performed under general anesthetic and multiple biopsy specimens are obtained. If no tumor is detected, the frequency of follow-up visits is reduced.
If tumor should be detected either macroscopically or histologically, then further treatment of the residual tumor consisting of surgical removal, continued fluorouracil application, and necrotomy plus additional radical radiotherapy (70 Gy) is performed. Postoperative radiotherapy is also indicated in cases of regional metastases or when the tumor extends beyond the reach of the local chemotherapy, such as the cheek or infratemporal fossa. Likewise treatment of later local tumor recurrence following initial treatment success consists of surgical removal, frequent fluorouracil application, necrotomy, and radiotherapy.
Patients initially seen with regional lymph node metastases are treated additionally with a neck dissection and locoregional radiotherapy. Treatment of patients developing metastatic lymph nodes during follow-up depends on the situation with the local primary site: presence of local recurrence indicates standard treatment with neck dissection and full-dose radiotherapy; if there is no evidence of local recurrence, the patient undergoes a neck dissection followed by radiotherapy as indicated by the histological findings (ie, extranodal spread, >1 involved node or angioinvasion). Treatment of patients with distant metastases is tailored individually.
Between January 1976 and December 1997 seventy patients with adenocarcinoma of the ethmoid were referred for treatment to our department. Sixty-two of these patients were eligible for primary therapy according to the aforementioned regimen and form the basis of the following analysis (Table 1). Six patients (10%) were female; all others were male. Twenty-eight patients (45%) had a history of working with wood.
Eight patients were excluded for a variety reasons: 2 received no treatment (1 died before treatment was started and 1 remained untreated owing to severe Alzheimer disease from which the patient died). Six patients were treated by alternative methods: 2 patients received preoperative radiotherapy followed by debulking surgery alone, 2 patients had extensive intracranial extension requiring CFR (3%), and 2 patients received palliative therapy in the presence of metastatic disease.
According to the American Joint Committee on Cancer–International Union Aganst Cancer Classification,15 the tumors of 3 patients were classified as T1, 10 as T2, 24 as T3, and 25 as T4. Involvement of the anterior skull base was found in 40% (25 of 62 patients) and of the medial orbital wall and/or periorbital lining in 60% (37 of 62 patients) of patients. The orbit was preserved in all patients.
The minimum follow-up to date is 2 years (mean follow-up, 8.2 years). Overall survival and disease-free survival rates are shown in Figure 1. Disease-free survival was 96%, 87%, and 74% at 2, 5, and 10 years, respectively (Kaplan-Meier). Local and distant relapse-free survival are shown in Figure 2.
Thirty-two patients received preoperative low-dose radiotherapy during the pre-1988 phase of this study, while after 1988 thirty patients did not. There was no difference between the pre-1988 and post-1988 groups, 5-year disease-free survival being 84% and 89%, respectively, while 10-year survival was 74% and 74%, respectively.
None of these patients had residual tumor at inspection under anesthetic approximately 3 months after therapy. Eight patients (13%) developed a local tumor recurrence: 4 of them are alive and free of disease after 1 round of additional treatment with debulking surgery, local chemotherapy combined with a second necrotomy and radiotherapy, 1 is alive with disease after repeated local treatment, and 3 died of their tumor. One patient developed a regional metastasis and was salvaged with a radical neck dissection and survived for a further 7 years after surgery but then died of a second primary carcinoma of the lung. Five patients (8%) developed distant metastases; they all died of their disease.
The complications are listed in Table 2. Most complications were not serious. There were no postoperative deaths. Despite involvement of the anterior skull base in 25 patients, there was only 1 episode of postoperative meningitis that responded rapidly to antibiotic therapy. Five patients (8%) developed temporary postoperative cerebrospinal fluid leaks that were successfully managed with bed rest and local nasal tamponade with tetracycline-saturated gauze. None of these patients required further surgery for persistent leakage. Periorbital bruising and swelling was the most common postoperative complication occuring to varying degrees in many patients who had involvement of the lamina papyracea (25 patients [40%]). This was of minor concern to the patients and generally resolved spontaneously within 1 week of surgery. There were no episodes of visual disturbance or of intraorbital bleeding. At the conclusion of this study, 42 patients were alive and free of disease; 9 have died of their disease, 2 are alive with disease, and 10 died of other causes without signs of locoregional or distant disease.
This treatment regimen seems to improve the prognosis of patients with adenocarcinoma of the ethmoidal sinuses. In animal experiments it has been demonstrated that local chemotherapy causes not only necrosis of tumor cells but also initiates a strong potentiation of the immune system.16- 19 This is one possible explanation for the apparent success of this form of treatment. It clearly goes against the concept of radical excisional surgery for these tumors20 but is perhaps more in line with a philosophy of organ preservation surgery.
In 1985 we published our results of this form of treatment for all histological types of paranasal sinus carcinoma.12 Our results then were similarly good not only for adenocarcinoma (100% 5-year survival, N = 20) but also were very encouraging for squamous cell carcinoma and undifferentiated carcinoma (50% 5-year survival). It remains unclear to us why in the past 15 years this form of treatment has not been more widely accepted for its simplicity but more importantly for its higher success rates when compared with other forms of conventional therapy.
In 3 recent reviews3,4,7 the long-term results of CFR for adenocarcinoma of the ethmoids have been published (Table 3). The 5-year survival rates vary from 39% to 57%. Comparing CFR with debulking surgery and topical chemotherapy, the results of CFR seem inferior.
If the validity of a treatment regimen is based on the clinical outcome, eg, survival data, then our current form of therapy would seem to be the treatment of choice for these tumors. We acknowledge that this hypothesis has not been tested in a clinical controlled trial. We suggest, however, that such a trial would be difficult to administer in view of the rarity of these tumors and the differences that exist between the forms of treatment. We believe that our experience in a large tertiary referral center with a population base of about 3 million is likely to be representative of other published series.
We accept that CFR remains the treatment of choice for esthesioneuroblastoma as it provides to-date results that are superior to other forms of treatment.6 Adenocarcinoma of the ethmoids, however, is a different disease and, despite improvements in surgical and anesthetic techniques that have allowed CFR to be performed with minimal morbidity and mortality, survival data to date with this treatment remain disappointing. Our current choice of treatment, therefore, is surgical debulking followed by repeated topical fluorouracil therapy and necrotomy.
We have reported our 23-year experience with a surgical debulking and topical chemotherapy technique for the treatment of adenocarcinoma of the ethmoidal sinuses. To date our survival figures have been encouraging and seem superior to those of the more conventional form of treatment, namely, CFR. We would suggest that our form of treatment (surgical debulking plus repeated topical fluorouracil therapy and necrotectomy) should be more widely adopted as we believe it provides patients with a better chance of survival from this malignancy.
Accepted for publication November 10, 2000.
Corresponding author: Paul P. Knegt, MD, PhD, Department of Head and Neck Surgery, University Hospital Rotterdam Dijkzigt, 3000 CA Rotterdam, the Netherlands (e-mail: email@example.com).