Abstract
Objective. Explore the relationship between tumor location and prognosis among patients with Merkel cell carcinoma (MCC) of the head and neck.
Study Design. Retrospective analysis of large population database.
Setting. Surveillance Epidemiology and End Results (SEER) database/multiple settings.
Subjects and Methods. Using the SEER database, the authors identified patients with MCC of the head and neck. They compared clinicopathologic characteristics and disease-specific survival (DSS) between patients with head and neck MCC at different anatomic subsites. DSS was estimated by the Kaplan-Meier method, and a multivariable regression model was constructed so that independent predictors of DSS could be determined.
Results. Of the 2104 patients identified, 61.0% were men. The mean age at diagnosis was 77.5 years. Tumors were most commonly located on the face (61.1%). Scalp tumors were significantly larger (10.4% >5 cm, P = .0001) and more likely to present with distant metastasis (8.7%, P = .07) than other head and neck tumors. Lip tumors had the highest rate of invasion into bone, cartilage, and muscle (13.7%, P = .012), and ear tumors had the highest rate of nodal metastasis (63.2%, P = .011). Patients with scalp/neck and lip primary sites had significantly worse survival on univariable analysis (P = .0054 and P = .0007, respectively); however, the lip was the only site associated with worse survival on multivariable analysis (hazard ratio = 1.8, P = .005).
Conclusions. This study is the first to report lip location as an independent prognostic factor in head and neck MCC. More aggressive treatment of patients with MCC of the lip may be warranted to improve outcomes.
Merkel cell carcinoma (MCC) is a rare, aggressive cutaneous neoplasm of neuroendocrine origin. Accumulated exposure to ultraviolent B radiation and a tumor-associated Merkel cell polyomavirus likely increase the risk of developing MCC; hence, MCC most frequently affects elderly and immune-suppressed patients.1,2 Merkel cell carcinoma’s deceivingly indolent presentation as a benign-appearing, painless papule often delays its diagnosis, thus increasing the probability of regional and distant metastasis at presentation.1,3 Delayed treatment and a propensity for recurrence and metastasis contribute to MCC’s poor prognosis and high mortality rate.1,3
Merkel cell carcinoma most commonly occurs in the head and neck region, perhaps as a result of unprotected, chronic sun exposure.1,4 Previous studies that have analyzed the prognostic significance of anatomic site in other cutaneous malignancies have demonstrated that skin cancer of the head and neck may carry a less favorable prognosis compared with cutaneous malignancies in other regions.5-8 Furthermore, tumor location within the head and neck region is known to carry prognostic significance in both melanoma and nonmelanoma cutaneous malignancies.8-11 Tumor location has even been incorporated into the American Joint Committee on Cancer (AJCC) classification system for basal and squamous cell carcinomas. 12 However, the effect of tumor location on prognosis in MCC has yet to be elucidated. Exploration of the effect of head and neck anatomic subsite on prognosis is especially important in MCC because most of these tumors occur in the head and neck region. Therefore, our objective was to describe MCC by location within the head and neck and to determine the effect of primary site on prognosis in head and neck MCC.
Methods
Data Source
The Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute is a coordinated system of population-based cancer registries that is widely used to study a large variety of malignancies. Strategically located across the United States, the SEER registries routinely collect data on patient demographics, primary tumor site, tumor morphology, stage at diagnosis, first course of treatment, and follow-up vital status. The SEER program performs continuous quality control activities to ensure the collection of high-quality data, and the 17 registries included in our analysis represent an estimated 26% of the US population from several geographic regions. 13 This study was exempt from review by the institutional review board (IRB) at the Medical University of South Carolina.
Patient Selection
The SEER database was queried for patients diagnosed with cutaneous head and neck MCC based on the third edition of the International Classification of Diseases for Oncology (ICD-O-3) anatomic site and histology codes. Patient demographics, including age at diagnosis, sex, and race, as well as tumor factors were reviewed. Tumor factors included anatomic site of the primary tumor, tumor size and extension at presentation, and presence of lymph node and distant metastases. Primary site was categorized as (1) face, not otherwise specified (NOS); (2) scalp/neck; (3) skin of lip; (4) eyelid; and (5) external ear according to ICD-O-3 anatomic codes. Tumor size was analyzed according to the AJCC TNM classification system 12 as ≤2 cm, 2 to 5 cm, and >5 cm. Because of the variability in the SEER database of reporting clinical size as categorical or continuous variables, it was not possible to accurately analyze size as a continuous variable. Tumor extension was classified as (1) confined to the dermis, (2) through the entire dermis into subcutaneous tissue, and (3) through subcutaneous tissue into bone, cartilage, and skeletal muscle, as defined by SEER extent of disease and collaborative stage codes. Lymph node status was available for patients diagnosed after 1988 and was classified as pathologic node-negative, pathologic node-positive, and no pathologic lymph node examination (presumed clinical node-negative). Data were obtained from all 17 US cancer registries participating in the SEER program using SEER*Stat version 7.0.4 (www.seerstat.gov).
Statistical Analysis
Retrospective analysis was performed. We compared differences in clinicopathologic characteristics between patients with head and neck MCC at different anatomic subsites by 1-way analysis of variance (ANOVA), Kruskal-Wallis test, and χ2 testing where appropriate. Kaplan-Meier survival analysis was used to estimate 15-year disease-specific survival (DSS) for patients based on tumor location within the head and neck region. Survival time was calculated as the number of months between the date of diagnosis and the date of death up to 180 months or the date last known to be alive, whichever occurred first. Patients who were lost to follow-up or who survived beyond 180 months were coded as censored observations. The log-rank test assessed statistical differences between survival curves. The influence of age, sex, tumor size, extension, and the presence of regional and distant metastasis was determined by multivariate analysis using the Cox proportional hazards model. Hazard ratios (HRs) with 95% confidence intervals were calculated to estimate the risk of death. A probability value of less than .05 was considered significant. Stata version 10.0 (StataCorp, College Station, Texas) statistical software was used for all statistical analysis.
Results
Study Population
We identified 2104 patients in the SEER database who were diagnosed with cutaneous MCC of the head and neck between 1980 and 2008. Tumors were most commonly located on the face (61.1%), followed by the scalp/neck (21.2%), ear (7.0%), eyelid (5.4%), and lip (5.3%). Most patients were white men older than age 75 years. Patient characteristics according to tumor location within the head and neck region are summarized in Table 1 . Patients with scalp/neck tumors were the youngest, and patients with lip tumors were the oldest (mean age 76.6 and 79.9 years, respectively; P = .041). Eyelid tumors were slightly more common in women (54.4%); however, tumors in other head and neck locations occurred more frequently in men. Tumors located on the ear were associated with the highest percentage of male patients (77.6%).
Patients with Merkel Cell Carcinoma of the Head and Neck: Demographics
Abbreviation: NOS, not otherwise specified.
One-way analysis of variance.
χ2 test.
Kruskal-Wallis test.
Tumor Characteristics
Tumor characteristics are summarized by head and neck subsite in Table 2 . Most head and neck tumors were ≤2 cm in size (68.2%). Lip tumors were the smallest (89.8% ≤2 cm), and tumors on the scalp/neck (10.4% >5 cm) were the largest (P < .001). Most head and neck tumors were limited to the dermis (76.4%); however, tumors on the lip were significantly more likely to invade skeletal muscle, bone, and cartilage (13.7%, P = .012). Eyelid tumors were the least invasive (86.0% limited to dermis, P = .012).
Patients with Merkel Cell Carcinoma of the Head and Neck: Tumor Characteristics
Abbreviation: NOS, not otherwise specified.
Tumor size not recorded for 1050 patients (48.7% face, 48.8% scalp/neck, 45.1% lip, 53.1% eyelid, 45.9% ear).
Kruskal-Wallis test.
Tumor extension not recorded for 322 patients (9.1% face, 13.1% scalp/neck, 12.0% lip, 10.6% eyelid, 7.1% ear).
χ2 test.
Regional and Distant Metastasis
Table 3 displays data describing lymph node metastasis by anatomic subsite. Overall, 25.7% of patients had lymph nodes histopathologically examined. Among those, positive nodes were identified in 50.1%. Tumors located on the ear and eyelid had the highest and lowest rate of lymph node positivity, respectively (63.2% and 30.4%, P = .011).
Patients with Merkel Cell Carcinoma of the Head and Neck: Nodal Metastasis
Abbreviation: NOS, not otherwise specified.
χ2 test.
The incidence of distant metastasis at presentation was 5.7% in head and neck MCC overall ( Table 4 ). Although the frequency of distant metastasis at presentation was not significantly different between subsites, there was a trend toward an increased rate of distant metastasis in scalp/neck tumors (P = .07).
Patients with Merkel Cell Carcinoma of the Head and Neck: Distant Metastasis at Presentation
Abbreviation: NOS, not otherwise specified.
Kruskal-Wallis test.
Survival Analysis
Figure 1 illustrates the effect of tumor location within the head and neck region on DSS, and the results of multivariable Cox-proportional hazard regression analyses of clinicopathological factors with respect to DSS are shown in Table 5 . Univariable Kaplan-Meier analysis demonstrated that patients with tumors located on the lip and scalp/neck had significantly worse MCC-specific survival relative to patients with tumors located elsewhere in the head and neck region (P = .0007 and .0054, respectively). On multivariable analysis controlling for age at diagnosis, sex, tumor size and extension, nodal metastasis, and distant metastasis, lip was the only anatomic site independently associated with an increased risk of death from MCC (HR = 1.8, P = .005).
Merkel cell carcinoma–specific survival by head and neck subsite. Overall, P = .0012. Lip vs face, P = .0007. Scalp vs face, P = .0054. Lip vs scalp, P = .16. Face vs eyelid vs external ear, P = .72. NOS, not otherwise specified.
Multivariable Analysis of Prognostic Factors in Merkel Cell Carcinoma of the Head and Neck
Abbreviations: CI, confidence interval; HR, hazard ratio; NOS, not otherwise specified.
P < .05.
Discussion
Tumor location within the head and neck region carries prognostic significance in a number of cutaneous malignancies, including melanoma, which is an aggressive neuroendorine-derived cutaneous neoplasm that is often considered similar to MCC.8-11 However, the effect of tumor location on prognosis in MCC has not been reported previously. To our knowledge, we are the first to report differences in tumor characteristics and survival with respect to location in head and neck MCC.
Several studies have demonstrated that MCC occurs most frequently in the head and neck region and in males. 1 In concordance, most patients that we identified with head and neck MCC overall were men. Male predominance was most notable in scalp/neck and ear tumors. However, sex distribution among face and eyelid tumors was nearly even. Merkel cell carcinoma’s pattern of sex-related occurrence by location demonstrates the role of ultraviolet radiation exposure in the pathogenesis of MCC. These data suggest that the frequently reported increased incidence of MCC in males may be an effect of ultraviolet radiation exposure.
Previous studies have reported that tumor size is one of the most important factors related to prognosis in MCC. 14 Accordingly, size is used to stage MCC in the current AJCC classification system. 12 However, it is interesting that despite harboring significantly smaller tumors than other head and neck sites, the lip was associated with worse survival on both univariable and multivariable analysis. Furthermore, multivariable analysis demonstrated that tumor location significantly affected survival; however, survival was not significantly different between patients with tumors ≤2 cm and those with tumors >5 cm. Thus, it appears that tumor location, rather than tumor size, may be a more accurate predictor of survival in head and neck MCC.
Distant tumor invasion into skeletal muscle, cartilage, and bone, which is an established poor prognostic factor in MCC, occurred most frequently in the lip. Still, when controlling for tumor invasion in multivariable analysis, the lip remained independently associated with poorer survival. Even so, the significantly higher rate of distant tumor invasion in lip tumors further demonstrates the more aggressive nature of tumors in this location.
Previous studies have also identified the lip as a high-risk primary site for cutaneous malignancies.15-17 The poor prognosis that is associated with lip tumors may be a consequence of its midline location and bilateral lymphatic drainage, which provides an additional route for tumor metastasis. However, our data clearly demonstrate that other factors are inherent here, given the lip’s lower rate of nodal and distant metastasis relative to other head and neck sites. The lip’s proximity to the submucosa and its dense neurovascular supply possibly allows for easier and faster invasion as well. Regardless, more aggressive treatment, even for early stage lip tumors, is likely warranted to improve outcomes for these patients.
Scalp/neck tumors were notable for having the largest tumor size and highest rate of distant metastasis at presentation. As a result, patients with scalp/neck tumors had significantly worse survival on univariable analysis. However, scalp/neck location alone was not independently associated with survival on multivariable analysis; thus, earlier detection would likely improve outcomes for these patients. Our data emphasize the importance of thorough examination of the scalp and neck skin during screening and routine physical examinations, especially in elderly and immune-compromised patients and those with a history of previous skin cancers who are most at risk for MCC.
As previous studies have demonstrated, we identified a high rate of nodal metastasis in head and neck MCC. 1 In concordance with other skin cancer studies, this rate was especially high for tumors located on the ear, perhaps because of the dense concentration of anterior and posterior cervical lymph nodes in this area.18-20 The lowest rate of node positivity, 30.4% in eyelid tumors, is still significantly high and supports treatment guidelines that recommend routine treatment of regional lymph nodes at presentation regardless of clinical node status and tumor location. 21 However, it is interesting that despite the substantial difference in the incidence of nodal metastasis between eyelid and ear tumors (30.4% vs 63.2%), survival was remarkably similar among patients with tumors at each location (P = .45).
Conclusion
Tumor location is an important prognostic factor in head and neck MCC. We are the first to report lip primary site as an independent predictor of worse survival in head and neck MCC. More aggressive treatment of patients with MCC of the lip may be warranted to improve outcomes.
Tumors on the scalp/neck also carry a worse prognosis relative to tumors located elsewhere in the head and neck region; however, this is likely a result of delayed presentation and treatment. It is important that otolaryngologists who are involved in the ongoing care and surveillance of patients with other head and neck cutaneous malignancies who are at an increased risk for developing MCC, as well as other cutaneous malignancies, perform a thorough screening examination that includes careful examination of the skin of the neck and scalp at each encounter because earlier recognition may improve survival, especially for those with scalp/neck tumors.
Author Contributions
Disclosures
Footnotes
No sponsorships or competing interests have been disclosed for this article.