Abstract
Background
The purpose of this study is to present our experience treating patients with squamous cell carcinoma (SCC) from an unknown head and neck primary site and to determine whether a policy change eliminating the larynx and hypopharynx from the radiotherapy (RT) portals has impacted outcome.
Methods
One hundred seventy-nine patients received definitive RT with or without a neck dissection for SCC from an unknown head and neck primary site. RT was delivered to the ipsilateral neck alone or both sides of the neck and, usually, the potential mucosal primary sites. The median mucosal dose was 5670 cGy. The median neck dose was 6500 cGy. One hundred nine patients (61%) received a planned neck dissection.
Results
Mucosal control at 5 years was 92%. The mucosal control rate in patients with RT limited to the nasopharynx and oropharynx was 100%. The 5-year neck-control rates were as follows: N 1 , 94%; N 2a , 98%; N 2b , 86%; N 2c , 86%; N 3 , 57%; and overall, 81%. The 5-year cause-specific survival rates were as follows: N 1 , 94%; N 2a , 88%; N 2b , 82%; N 2c , 71%; N 3 , 48%; and overall, 73%. The 5-year overall survival rates were as follows: N 1 , 50%; N 2a , 70%; N 2b , 59%; N 2c , 45%; N 3 , 34%; and overall, 52%. Eleven patients (7%) developed severe complications.
Conclusion
RT alone or combined with neck dissection results in a high probability of cure with a low risk of severe complications. Eliminating the larynx and hypopharynx from the RT portals did not compromise outcome and likely reduces treatment toxicity.
1
Introduction
Approximately 3% of patients with squamous cell carcinoma (SCC) of the head and neck present with metastatic cervical adenopathy and an unknown primary site. Patients with adenopathy primarily in the upper neck (levels 2 and 3) have a good prognosis after treatment, whereas those with a low-neck presentation (levels 4 and/or supraclavicular fossa) commonly arise from a primary site below the clavicles and have a poor prognosis. The location of the occult primary tumor, if detected, probably varies geographically. Patients in areas where nasopharyngeal carcinoma is endemic are more likely to have an occult nasopharyngeal cancer. In contrast, the vast majority of occult primary cancers detected in the United States are found in the tonsillar fossa and the base of the tongue .
The management of patients with SCC of the head and neck from an unknown primary site is controversial, and varies from treatment of the neck alone to elective radiotherapy (RT) to the potential mucosal primary sites (nasopharynx, oropharynx, supraglottic larynx, and hypopharynx) and both sides of the neck . The treatment philosophy at our institutions has been similar in the periods included in this study. The goal of this article is to present our treatment outcomes with this entity and discuss the optimal management of such patients. Our treatment philosophy changed in 1997; and we have since irradiated the oropharynx, nasopharynx, and both sides of the neck, eliminating RT to the larynx and hypopharynx . One of the goals of this study is to determine whether this policy change resulted in an increased risk of out-of-field mucosal failures.
2
Methods and materials
Patients with SCC from an unknown head and neck primary site treated with curative intent at the University of Florida (139 patients) and University of Wisconsin (40 patients) were reviewed under the approval of the institutional review boards. Patients treated at the University of Florida began therapy between November 1964 and April 2005, and those managed at the University of Wisconsin were treated between October 1990 and September 2006. One-hundred fifty-seven patients were male (88%); the median age was 61 years (range, 26 to >89 years). Median follow-up was 4.2 years (range, 0.2–25.4 years). Median follow-up for survivors was 6.8 years (range, 1.1–23.4 years). Patients were staged according to the 2002 American Joint Committee on Cancer staging system . The SCC was poorly differentiated in 85 patients (47%), and well or moderately differentiated or not otherwise specified in the remainder. The diagnostic evaluation typically included obtaining a history and performing a physical examination (including a comprehensive head and neck examination by multiple examiners), chest radiography, direct laryngoscopy/panendoscopy and directed biopsies, and tonsillectomy in selected patients. Currently, a tonsillectomy is performed if sufficient lymphoid tissue is present in the tonsillar fossa to warrant the procedure. Computed tomography (CT) of the head and neck with or without magnetic resonance imaging was obtained in the vast majority of patients before direct laryngoscopy since approximately 1983 . Fluorodeoxyglucose (FDG)–single-photon emission CT or FDG–positron emission tomography was obtained in a subset of selected patients with equivocal findings on the radiographic evaluation. The N stage distribution and diagnostic evaluation are depicted in Table 1 . Thirty-two patients (18%) underwent a tonsillectomy: unilateral, 26 patients; bilateral, 6 patients.
| Parameter | No. of patients (%) |
|---|---|
| N stage | |
| N1 | 18 (10%) |
| N2a | 48 (27%) |
| N2b | 46 (26%) |
| N2c | 11 (6%) |
| N3 | 56 (31%) |
| Diagnostic evaluation | |
| CT | 132 (74%) |
| MRI | 15 (8%) |
| FDG-SPECT | 9 (5%) |
| FDG-PET or PET/CT | 16 (9%) |
| Direct laryngoscopy/panendoscopy | 156 (87%) |
| Tonsillectomy | 32 (18%) |
| Neck dissection | |
| None | 70 (39%) |
| Unilateral | 103 (58%) |
| Bilateral | 6 (3%) |
| RT volume | |
| Ipsilateral neck | 5 (3%) |
| OC/OPX | 7 (4%) |
| OPX/NPX | 28 (15%) |
| OPX/HPX | 1 (1%) |
| NPX, OPX, HPX | 138 (77%) |
All patients were treated with RT to volumes ranging from the ipsilateral neck alone to both sides of the neck and the potential mucosal primary sites extending from the nasopharynx to the hypopharynx ( Table 1 ). Since 1997, the mucosal sites included have been the oropharynx and nasopharynx because the majority of unknown primaries are likely to be in the tonsillar fossa and the base of the tongue; the larynx and hypopharynx have only been included for patients with level III nodal presentations. The small subset of patients who may have been selected for neck dissection alone without RT is not defined, and these patients were not included in the analysis. The median mucosal dose was 5670 cGy (range, 2400–7440 cGy); 4 additional patients received no mucosal dose. The median dose to the neck was 6500 cGy (range, 5000–8600 cGy). One-hundred sixty-eight patients (94%) were treated with once-daily RT, and 11 patients (6%) were treated with twice-daily fractionation. Thirteen patients received adjuvant chemotherapy: induction, 2 patients; concomitant, 11 patients. Planned neck dissection, when performed, usually was accomplished 4 to 6 weeks after RT. Planned neck dissection was performed in 109 patients (61%): unilateral, 103 patients; bilateral, 6 patients. Neck dissection was performed before RT in 44 patients and after RT in 65 patients.
All statistical computations were accomplished with SAS and JMP software (SAS Institute, Cary, NC). Estimates of local (mucosal) control, regional control, distant metastasis-free survival, cause-specific survival, and overall survival were attained with the Kaplan-Meier product-limit method . Multivariate analysis was accomplished with Cox regression; backward selection provided the most parsimonious final model from a group of selected explanatory variables for each end point . Variables included in the multivariate analysis were the following: N stage (N 1 –N 2 vs N 3 ), histologic differentiation (well or moderately differentiated or not otherwise specified vs poorly differentiated), neck dissection (none vs unilateral or bilateral), and mucosal RT (no vs yes). Severe complications were defined as those necessitating hospitalization or surgical intervention and/or resulting in death.
2
Methods and materials
Patients with SCC from an unknown head and neck primary site treated with curative intent at the University of Florida (139 patients) and University of Wisconsin (40 patients) were reviewed under the approval of the institutional review boards. Patients treated at the University of Florida began therapy between November 1964 and April 2005, and those managed at the University of Wisconsin were treated between October 1990 and September 2006. One-hundred fifty-seven patients were male (88%); the median age was 61 years (range, 26 to >89 years). Median follow-up was 4.2 years (range, 0.2–25.4 years). Median follow-up for survivors was 6.8 years (range, 1.1–23.4 years). Patients were staged according to the 2002 American Joint Committee on Cancer staging system . The SCC was poorly differentiated in 85 patients (47%), and well or moderately differentiated or not otherwise specified in the remainder. The diagnostic evaluation typically included obtaining a history and performing a physical examination (including a comprehensive head and neck examination by multiple examiners), chest radiography, direct laryngoscopy/panendoscopy and directed biopsies, and tonsillectomy in selected patients. Currently, a tonsillectomy is performed if sufficient lymphoid tissue is present in the tonsillar fossa to warrant the procedure. Computed tomography (CT) of the head and neck with or without magnetic resonance imaging was obtained in the vast majority of patients before direct laryngoscopy since approximately 1983 . Fluorodeoxyglucose (FDG)–single-photon emission CT or FDG–positron emission tomography was obtained in a subset of selected patients with equivocal findings on the radiographic evaluation. The N stage distribution and diagnostic evaluation are depicted in Table 1 . Thirty-two patients (18%) underwent a tonsillectomy: unilateral, 26 patients; bilateral, 6 patients.
