Spindle Cell Melanoma: Incidence and Survival, 1973‑2017
Total Page:16
File Type:pdf, Size:1020Kb
ONCOLOGY LETTERS 16: 5091-5099, 2018 Spindle cell melanoma: Incidence and survival, 1973‑2017 ZHE XU1*, PING SHI1*, FEILUORE YIBULAYIN2,5, LEI FENG2,3, HAO ZHANG3,4 and ALIMUJIANG WUSHOU2,3 1Department of Stomatology, Shenzhen Longhua District Central Hospital, Guangdong Medical University, Shenzhen, Guangdong 518110; Departments of 2Oral and Maxillofacial Surgery, 3Oral Biomedical Engineering Laboratory, and 4Epidemiology and Biostatistics, Shanghai Stomatological Hospital, Fudan University; 5Department of Preventive Medicine, School of Public Health, Fudan University, Shanghai 200001, P.R. China Received December 2, 2017; Accepted May 11, 2018 DOI: 10.3892/ol.2018.9247 Abstract. Spindle cell melanoma (SCM) is a rare morpholog- Introduction ical subtype of melanoma, which is relatively uncharacterized. The aim of the present study was to investigate the incidence Spindle cell melanoma (SCM) is a rare subtype of malignant of SCM, its general demographics, basic clinico-pathologic melanoma composed of spindled neoplastic cells arranged in features, treatment outcomes and disease‑specific prognostic sheets and fascicles (1). The diagnosis of SCM is challenging, factors. SCM cases were sampled from the Surveillance, as SCM may occur anywhere on the body and frequently Epidemiology and End Results (SEER) Program (1973-2017). mimics amelanotic lesions, including scarring and inflam- A total of 4761 SCM cases were identified, with a median age mation (2-4). Histologically, cytologic features of SCM are of 66 years. The female:male ratio was 0.62:1. Statistically indistinct and often confused with those of other epithelial significant overall survival (OS) and disease‑specific survival neoplasms, including sarcomas and lymphomas (5-8). (DSS) rate differences were identified depending on age, sex, Immunohistochemistry is a helpful tool in distinguishing ethnicity, tumor location, T stage, N stage, M stage, patholog- SCM from other sarcomas and carcinomas (9,10). However, ical grade, AJCC stage, SEER stages and surgical treatment diagnosis remains a challenge as a number of sarcomas share (P<0.05). Multivariate Cox regression analysis revealed that some morphological and immunohistochemical features with age >66 years, T3+T4 stage disease, positive N stage and SCM (5,11). Differentiation of SCM from desmoplastic mela- SEER historic stage of regional and distant metastasis tumor noma is difficult because both melanomas are characterized by were associated with poor DSS and OS rates. In summary, atypical, spindled, malignant melanocytes. However, the size SCM was most common in Caucasian people of 60~80 years of spindle cell collagen areas and the immunohistochemical of age with a predominance in males. Patient's age, ethnicity, markers, S100, MelanA and Tyrosinase, allow differential T stage, N stage, and SEER historic stage were identified diagnosis (10). Therefore, the integration of clinical and histo- as independent prognostic factors of SCM in terms of DSS logical assessment is essential for the diagnosis of SCM (2,8). and OS. Diagnosis of SCM is often delayed until patients exhibit advanced-stage disease, typically with widespread metastasis and poor treatment outcomes (3,6,12,13). A limited number of case reports and incomplete retro- spective case studies of the differential diagnostic viewpoints of SCM exist (3,4,11‑15). To the best of our knowledge, few studies have reported SCM incidence, clinicopathologic Correspondence to: Dr Alimujiang Wushou, Department of Oral features, treatment, treatment outcome and disease‑specific and Maxillofacial Surgery, Shanghai Stomatological Hospital, Fudan independent prognostic factors. Thus, the present study University, 356 Beijing East Road, Shanghai 200001, P.R. China performed a retrospective analysis of a series of clinical cases E-mail: [email protected] using data from the Surveillance, Epidemiology and End Results (SEER) Program. *Contributed equally Materials and methods Abbreviations: SCM, spindle cell melanoma; OS, overall survival; DSS, disease specific survival; SEER, the surveillance, epidemiology and end results; AJCC stage, American Joint Committee on Cancer Data collection. In the present study, data was analyzed from stage; HR, hazard ratio; CI, confidence interval the SEER Program, National Cancer Institute Public Use Dataset, which contains publically available records of 18 Key words: spindle cell melanoma, incidence, prognostic factor, the population-based cancer registries, which together represent surveillance, epidemiology and end results 28% of the USA population. Data were extracted regarding patients with a primary diagnosis of SCM, according to the International Classification of Diseases for Oncology, Third 5092 XU et al: SPINDLE CELL MELANOMA Edition (ICD-O-3), using histology codes: 8772/3 (16). Cases were excluded if treatment or outcome data were unavailable for survival analysis. The data extraction was carried out with the official software SEER*Stat, version 8.3.4. (URL: https://seer.cancer.gov/data/). Statistical analyses. Overall Statistical analysis was accom- plished using Statistical Package for Social Sciences (SPSS; version 23.0, for Windows; IBM Corp., Armonk, IL, USA). χ2 test or Fisher's exact test was used to analyze associations among baseline parameters. The primary endpoint in the present study was considered to be the date of SCM-associated mortality. The time point between the date of diagnosis and the date of SCM-associated mortality was defined as disease‑specific survival (DSS). Mortalities associated with SCM were considered to be events, while deaths attributed to Figure 1. The age and sex distribution of spindle cell melanoma cases. other causes were considered to be ‘censored observations’. In terms of overall survival (OS) and DSS rates, Kaplan-Meier method and the log-rank test were utilized and multivariate metastasis were factors independently associated with worse Cox proportional hazard models were used to identify signifi- OS (Table III). cant risk factors for survival outcomes. All statistical tests Significant differences in the DSS analysis were also were two-sided, and P<0.05 was considered to indicate a identified depending on age (P<0.001), sex (P<0.001), statistically significant difference. pathological grade (P<0.001), AJCC stage (P<0.001), ethnicity (P=0.009), T stage (P<0.001), tumor location (P<0.001), Results M stage (P<0.001), treatment modalities (P<0.001), and SEER historic stage (P<0.001) (Fig. 3). Univariate Cox regression Study population and characteristics. The following demo- analysis demonstrated that age, ethnicity, sex, tumor location, graphic and clinicopathological characteristics were selected pathological grade, AJCC stage, T stage, N stage, M stage, for analysis: Age at diagnosis, ethnicity, primary tumor loca- SEER historic stage and treatment modalities were associated tion, Tumor-Node-Metastasis (TNM) stage, American Joint with DSS (Table II). The multivariate Cox regression model Committee on Cancer (AJCC) stage, pathological grade (the revealed that age >66 years T3+T4 stage, positive N-stage and American Joint Committee on Cancer/Union for International SEER historic stage of regional and distant metastasis were Cancer Control staging system), SEER historic stage, treatment independently associated with a poor OS rate (Table III). modalities, vital status and follow-up time (17). Unfortunately, complete data was not available for all cases. Discussion Data from 4,761 patient diagnosed with SCM between 1973 and 2017 was retrieved from the SEER database. The As a morphological variant of melanoma, SCM is rare and total cohort consisted of 1,829 women and 2,932 men, with its incidence has been variably reported between 3 and a female:male ratio of 0.62:1. The patients' age ranged from 14% ���������������������������������������������������of all melanoma cases������������������������������ (including desmoplastic mela- 3‑101 years and a median age of 66 years. The age and sex noma) (15,18,19). Diagnosis of SCM is challenging and distributions are presented in Fig. 1. The median follow-up time awareness of its clinical and cytological features as well as was 53 months (range, 0-500 months). Regarding ethnicity, immunohistochemical markers are essential to reach the Caucasian people accounted for 96.7% of the study popula- correct diagnosis (9,10,15). Due to the rarity of SCM, its tion. The majority of the cases of SCM had originated from clinical and prognostic characteristics remain to be fully the skin, and the eye and bony orbits were the second-most elucidated. To the best of our knowledge, the present study affected tumor site. Surgical resection was performed in 88.7% is the first to investigate SCM incidence as well as survival cases. The basic demographic and clinicopathologic char- analysis on a large scale. acteristics of the whole patients are summarized in Table I. The present study demonstrates that the incidence of SCM Kaplan-Meier analysis was utilized for time-to-event analysis. was highest in the 6-8th decade of life in males. Caucasian Statistically significant differences in OS rate were identified people accounted for the majority of the study population. depending on age (P<0.001), sex (P<0.001), tumor location SCM lesions originated most commonly from the skin and (P<0.001), ethnicity (P=0.008), AJCC stage (P<0.001), T stage eyes, and the bony orbits were the second-most affected tumor (P<0.001), pathological grade (P<0.001), N stage (P<0.001), site. SCM shares