The natural history and patterns of metastases from mucosal melanoma: an analysis of 706 prospectively-followed patients

Transcription

1 Annals of Oncology 28: , 2017 doi: /annonc/mdw694 Published online 30 December 2016 ORIGINAL ARTICLE The natural history and patterns of metastases from mucosal melanoma: an analysis of 706 prospectively-followed patients B. Lian 1,C.L.Cui 1, L. Zhou 1, X. Song 2, X. S. Zhang 3,D.Wu 4,L.Si 1, Z. H. Chi 1, X. N. Sheng 1,L.L.Mao 1, X. Wang 1, B. X. Tang 1, X. Q. Yan 1, Y. Kong 1,J.Dai 1,S.M.Li 1, X. Bai 1, N. Zheng 5, C. M. Balch 6 & J. Guo 1 * 1 Department of Renal Cancer & Melanoma, Peking University Cancer Hospital & Institute, Beijing; 2 Department of Melanoma, Yunnan Cancer Hospital, Kunming; 3 Department of Melanoma, SUN YAT-SEN University Cancer Center, Guangzhou; 4 Department of Melanoma, The First Hospital of Jilin University, Changchun; 5 Clinical Pharmacology Research Centre, Peking University Cancer Hospital & Institute, Beijing, China; 6 Department of Surgical Oncology, University of Texas MD Anderson Cancer Center, Houston, Texas, USA *Correspondence to: Prof. Jun Guo, Key laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Department of Renal Cancer and Melanoma, Peking University Cancer Hospital & Institute, 52# Fucheng Road, Haidian District, Beijing, China, Tel: þ ; Fax: þ ; guoj307@126.com The first two authors contributed equally to this work. Background: We examined whether mucosal melanomas are different in their clinical course and patterns of metastases when arising from different anatomic sites. Our hypothesis was that metastatic behavior would differ from primary mucosal melanomas at different anatomical sites. Patients and methods: Clinical and pathological data from 706 patients were compared for their stage distribution, patterns of metastases, CKIT/BRAF mutation status, and overall survival for different anatomical sites. Results: The anatomic sites of the primary mucosal melanomas were from the lower GI tract (26.5%), nasal cavity and paranasal sinuses (23%), gynecological sites (22.5%), oral cavity (15%), urological sites (5%), upper GI tract (5%), and other sites (3.0%). At initial diagnosis, 14.5% were stage I disease, 41% Stage II, 21.5% Stage III, and 23.0% stage IV. Predominant metastatic sites were regional lymph nodes (21.5%), lung (21%), liver (18.5%), and distant nodes (9%). Oral cavity mucosal melanoma had a higher incidence of regional nodal metastases (31.7% versus 19.8%, P ¼ 0.009), and a higher incidence of lung metastases (32.5% versus 18.5%, P ¼ 0.007) compared to other primary mucosal melanomas. There was a 10% incidence of CKIT mutation and 12% BRAF mutation. Mucosal melanomas from nasal pharyngeal and oral, gastrointestinal, gynecological, and urological had a similar survival with a 1-year survival rate (88%, 83%, 86%), 2-year survival rate (66%, 57%, 61%), 5-year survival rate (27%, 16%, 20%), respectively. Conclusions: The largest sample size allows, for the first time, a comparison of primary melanoma stage and patterns of metastases across anatomical sites. With few exceptions, the presenting stages, incidence of nodal and distant metastases, the site of predilection of distant metastases, or overall survival were similar despite different primary anatomic sites. These findings suggest that clinical trials involving mucosal melanomas and the administration of systemic therapy can be applied equally to mucosal melanomas regardless of their primary anatomic site. Key words: mucosal melanoma, natural history, patterns of metastases, overall survival Introduction Mucosal melanoma is a subtype of melanoma that originates from melanocytes in mucosal membranes. It is aggressive with five overall survival ranging from 22% to 34%, depending in part upon the distribution of specific sites of primary mucosal melanomas in different series [1 5]. Depending on different primary locations, mucosal melanomas are usually classified as Head and Neck (Nasal pharyngeal and Oral), Gastrointestinal (Upper GI and Lower GI), Gynecological, Urological, and Other VC The Author Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please journals.permissions@oup.com.

2 Annals of Oncology sites. Previous research has demonstrated that mucosal melanomas are distinctive from cutaneous melanoma with regard to its clinical course and prognosis [4, 6]. There have been few epidemiologic and natural history studies about mucosal melanoma, primarily because of its extremely low incidence in most countries. But in Asia, it is the second common subtype after acral lentiginous melanoma (ALM), with a percentage incidence of 22 25% [2, 7]. Peking University Cancer Hospital & Institute serves as a national referral center in China for melanoma patients and therefore has a rich experience treating melanoma patients who are seen uncommonly among a predominately Caucasian population in the Western world (i.e.: United States, Europe, and Australia). The objective of this study is to analyze the natural history and patterns of metastases of mucosal melanoma from different primary sites. The hypothesis we examined was whether mucosal melanomas are different in their clinical course when arising from diverse anatomic sites, or whether they are similar in their natural history and patterns of metastases. This information is important when planning treatment approaches for mucosal melanoma patients and in the design of clinical trials involving systemic therapies. Description of methods and statistics Database design After the approval of institutional review board of the Peking University Cancer Hospital & Institute. A database was prospectively designed to identify all patients with mucosal melanoma from Peking University Cancer Hospital & Institute, Yunnan Cancer Hospital, SUN YAT-SEN University Cancer Center and the First Hospital of Jilin University. A total of 706 patients referred between December 2005 and December 2013 were involved in this study. We excluded patients with vulvar and penile melanoma, since these are regarded as an uncommon presentation of cutaneous melanoma. This is a collaborative study with other Centers contributing data as follows: Peking University Cancer Hospital & Institute, Beijing, China, 595 patients (84% of the total series); Yunnan Cancer Hospital, Kunming, China, 55 patients (8%); SUN YAT-SEN University Cancer Center, Guangzhou, China, 30 patients (4%); The First Hospital of Jilin University, Changchun, China, 26 patients (4%). Definitions of clinical, anatomical and pathological criteria. The characteristics of the patient and their pathology entered into the database included: patient age, sex, performance score, anatomic location, depth of invasion, ulceration, initial stage, lymph node metastasis, distant metastasis, elevation of serum LDH, CKIT mutation, BRAF mutation of all patients presented with pathologically confirmed mucosal melanoma. Anatomic sites were defined as follows: head and neck (including nasal cavity, paranasal sinuses, and oral cavity), upper gastrointestinal (including esophagus, gastric, small bowel), lower GI (including colorectal and anus), gynecological (including vaginal, cervix, and uterine), urological (bladder, ureter, and kidney) plus other primary sites of mucosal origin. Definitions of pathological criteria. All patient pathology was entered into the database had a central review at Peking University Cancer Hospital & Institute for presence or absence of tumor ulceration, tumor thickness, depth of invasion, histological type. The stage was modified based on AJCC staging system Seventh Edition [8]. Our definition of TNM staging for mucosal melanoma was defined as following: T1 for tumor invading submucosa; T2 for tumor invading the muscularis propria; T3 for tumor invading adventitia; T4 for tumor invading adjacent structures; N1 for 1 regional metastatic node; N2 for 2 3 regional metastatic nodes; N3 for 4 or more regional metastatic nodes; and M1 for distant metastases [9]. These definitions are similar to that used by others [10]. Results Original article Overall results for 706 patients from all anatomical sites There were 706 patients with mucosal melanoma in this study (Table 1). The median age at diagnosis was 55 years (81% between 40 and 70 years), and 63% were female. The anatomic sites of the primary melanomas were: 26.5% (188 patients) arising from the lower GI tract, 23% (164 patients) from the nasal cavity and paranasal sinuses, 22.5% (159 patients) from gynecological sites, 15% (104 patients) from the oral cavity, 5% (35 patients) from urological sites, 5% (33 patients) from the upper GI tract, and 3% (23 patients) from other sites. The depth of tumor invasion of the primary melanomas were categorized as T1, T2, T3, and T4 melanomas, which was distributed as 8% T1 melanomas, 24% as T2, 42% as T3, and 26% as T4. The majority (59%) of tumors had an ulcerated surface on pathological examinations. At the initial diagnosis, 14.5% (100 patients) were diagnosed as stage I disease, 41% (291 patients) as Stage II, 21.5% (152 patients) as Stage III, and 23% (163 patients) as stage IV. Among Stage III patients with lymph node metastases, there were 31% with N1 stage, 30% with N2 stage, and 39% with N3 stage. The analysis of stage IV patients included those who were diagnosed with stage IV initially and patients who had progressed from stage I III when we calculated the distribution of distant metastases (totally 547 patients). For Stage IV patients, 49(9.0%) presented initially with distant lymph node metastasis, 101(18.5%) had liver only metastases, 116(21.0%) had lung only, 39(7.0%) had combined liver and lung, while 80(14.5%) patients had metastases at a variety of other sites, and an elevation of serum LDH was observed in 162 patients (30.0%). A minority of patients had a CKIT mutation (70 patients or 10%) or a BRAF mutation (85 patients or 12.0%). Characteristics of mucosal melanomas from specific anatomic sites When examining separate anatomic sites, they were similar in metastatic distribution when comparing data across presenting stages of disease, incidence of nodal and distant metastases, or the site of predilection of distant metastases. Also there was not much difference with regard to CKIT mutation and BRAF mutations (Table 2). Patterns of metastases based upon presenting clinical stage of disease Clinically localized mucosal melanoma (stage I and II). There were 391 patients who presented with localized mucosal melanoma (Table 3).The most common primary sites were nasal cavity and paranasal sinuses (26.5%), gynecologic (25%), lower GI (23%), and oral cavity (14%). Most of these lesions were still Volume 28 Issue doi: /annonc/mdw

3 Original article Table 1. Clinical and pathological characteristics of all 706 mucosal melanoma patients All mucosal melanoma 706 pts (%) Age (years) Median (range) 55 (17,86) < > Male/female 37/63 Anatomic site Nasal cavity and sinuses 23.0 Oral cavity 15.0 Upper GI 5.0 Lower GI 26.5 Gynecologic 22.5 Urologic 5.0 Other sites 3.0 Depth of invasion T1 8.0 T T T Ulceration present Absent 34.0 Present 59.0 Not available 7.0 Initial stage Stage I 14.5 Stage II 41.0 Stage III 21.5 Stage IV 23.0 Lymph node metastases (stage I III) NO 72.0 N1 8.5 N2 8.5 N Distant metastases Distant lymph node 9.0 Liver only 18.5 Single 13.0 Multiple 87.0 Lung only 21.0 Single 21.5 Multiple 78.5 Liver and lung 7.0 Other sites 14.5 Elevation of serum LDH 30.0 ckit mutation Present 10.0 BRAF mutation Present 12.0 Performance status (%) locally advanced (74.5% were T3 or T4 lesions), including an ulcerated surface in 59%. Only 19% had an elevated serum LDH level. Mucosal melanomas with metastases to regional lymph node (stage III). There were 152 patients with regional nodal metastases arising from their mucosal melanoma primary site (Table 3). The most common primary sites were lower GI (31.5%), oral Annals of Oncology cavity (22%), nasal cavity and paranasal sinuses (20%), and gynecologic (15%). Most of these lesions were still locally advanced (71% were T3 or T4 lesions), including an ulcerated surface in 56%. Only 14% had an elevated serum LDH level. Mucosal melanomas presenting with distant metastases (stage IV). There were 163 patients who presented with distant metastases (Table 3). The most common primary sites were lower GI (30%), gynecologic (24.5%), nasal cavity and paranasal sinuses (18.5%), and oral cavity (9.5%). Most of these lesions were still locally advanced (49.5% were T3 or T4 lesions), including an ulcerated surface in 62%. A greater percentage of patients (42%) had an elevated serum LDH level. Patterns of metastases from primary mucosal melanomas Regional lymph nodes. There were 152 patients (21.5% of the total) with regional lymph node metastases. Of these, 47 patients had N1, 46 had N2, and 59 had N3 stage. Mucosal melanomas arising from the oral cavity had a higher incidence of regional nodal disease compared to other sites with mucosal melanoma (31.7% versus 19.8%, P ¼ 0.009). Liver metastases. There were 101 patients (18.5%) with metastases confined to liver only. Only 13% of these patients had an isolated metastasis. Mucosal melanoma from urological sites had a higher incidence of liver only metastases compared to other sites (31.0% versus 17.1%), but the difference was not statistically significant (P ¼ 0.085). Lung metastases. There were 116 patients (21.0%) with metastases confined to the lung. Only 21.5% of these patients had an isolated metastasis. Mucosal melanomas from the oral cavity had a higher incidence of lung only metastases compared to other sites (32.5% versus 18.5%), a difference that was statistically significant (P ¼ 0.007). Combined liver and lung metastases. There were 39 patients (7.0%) with combined liver and lung metastases. Other sites. There were 49 patients (9.0%) with metastases in distant lymph nodes. Also, there were 80 patients (14.5%) with metastases at a variety of other sites. Comparisons of metastatic incidence from different primary sites. Comparing the incidence of metastatic disease at different sites, we found only a few statistically significant differences: 1. Oral cavity mucosal melanoma had a higher incidence of regional nodal disease compared to metastatic sites arising from other primary mucosal melanoma (31.7% versus 19.8%, P ¼ 0.009). 2. Oral cavity mucosal melanoma had a higher incidence of distant lung metastases compared to metastases arising from other primary mucosal melanoma (32.5% versus 18.5%, P ¼ 0.007), and 3. Urologic mucosal melanoma had a higher incidence of liver metastases only than other primary mucosal melanomas (31.0% versus 17.1%), but the sample size was small and the difference was not statistically significant (P ¼ 0.085). 870 Lian et al. Volume 28 Issue

4 Annals of Oncology Table 2. Clinical and pathological characteristics of mucosal melanomas at individual anatomic sites Factors Nasal cavities %(164 pts) Oral cavities %(104 pts) Upper GI %(33 pts) Lower GI %(188 pts) Original article Gynecologic %(159 pts) Urologic %(35 pts) Other sites a %(23 pts) Age (years) Median (range) 56 (27,81) 55.5 (22,80) 52.5 (39,65) 59 (24,84) 52 (17,86) 57 (36,75) 46.5 (17,67) < > Male/Female 50/50 44/ / /51.0 0/ / /43.5 Depth of invasion T T T T Ulceration NO YES NA Initial stage Stage I Stage II Stage III Stage IV Lymph node metastases (stage I-III) Distant metastases Comparisons of overall survival from different primary sites. Mucosal melanomas from nasal pharyngeal and oral, gastrointestinal, gynecological and urological had a similar survival with a 1-year survival rate (88%, 83%, 86%), 2-year survival rate (66%, 57%, 61%), 5-year survival rate (27%, 16%, 20%), respectively. Discussion NO N N N Distant lymph node metastasis only Liver only Single Multiple Lung only Single Multiple Liver and lung Other sites Elevated LDH ckit mutation BRAF mutation KPS a There was a smaller number of patients who presented with mucosal melanomas at other primary sites, including respiratory (10 patients), gallbladder (3 patients), spinal meninges (6 patients), amygdala of the brain (2 patients), and prostate (2 patients). This is the largest published experience with mucosal melanoma, mainly from a single institution. The large sample size allows, for the first time, a comparison of primary melanoma presentation and patterns of metastases across anatomical sites. Our hypothesis was that metastatic behavior would differ from primary mucosal melanomas at different anatomical sites. However, this hypothesis could not be accepted even with this large data set, when comparing data across the presenting stages of disease, the incidence of nodal and distant metastases, nor the site of predilection of distant metastases. These findings suggest that the mutational events that result in mucosal melanomas are independent of the anatomic site where they arise, and further, that clinical trials involving mucosal Volume 28 Issue doi: /annonc/mdw

5 Original article Table 3. Clinical and pathological characteristics of all mucosal melanoma patients based upon presenting clinical stage of disease Factors Localized (Stage I and II) %(391 pts) Regional nodal metastases (Stage III) %(152 pts) Annals of Oncology Distant metastases (Stage IV) %(163 pts) Age (years) Median(range) 56 (17,86) 54.5 (26,80) 55 (25,80) < > Male/Female 36.0/ / /63.0 Location Nasal cavity Oral cavities Upper GI Lower GI Gynecologic Urologic Other Depth of invasion T T T T Ulceration NO YES NA Elevated LDH CKIT mutation BRAF mutation KPS melanomas and the administration of systemic therapy can be applied equally to mucosal melanomas regardless of their primary anatomic site. Of course, this hypothesis has to be proven at the genetic/molecular level with future studies. It is reported that mucosal melanoma seems have a generally older age distribution [4, 11]. The mean age of these patients was 55 years, which was similar with cutaneous melanoma patients in China [2]. The age distributions were similar among different anatomic locations. The male-to-female rate ratio (0.59) was essentially the same as that for cutaneous melanoma in American people ( % CI of ) [12]. This higher rate among women was influenced by the inclusion gynecologic tract female melanoma patient in our series (22.5%). Head and neck (nasal cavity, paranasal sinuses and oral cavities) was the most common primary site followed by lower GI and gynecologic anatomic sites, similar to the distribution reported from other countries [11]. More than half of the mucosal melanomas exhibited an ulceration of their mucosal lining (59.0%), a character which is an important and independent prognostic factor for cutaneous melanomas [2, 13]. Unlike the staging factors used for cutaneous melanoma, there was still no standardized thresholds of tumor staging mucosal melanomas using tumor thickness or depth of invasion [14]. For example, the median tumor thickness was 3.85 mm in our series of mucosal melanomas, more than twice that for cutaneous melanoma patients [12]. More than 44% of mucosal melanoma patients presented with metastatic disease at either regional sites (21.5%) or at distant locations (23.0%); in contrast to an incidence of only 14% among for 68,495 of patients with cutaneous melanoma diagnosed from 1992 to 2005 in America [15]. Both the incidence of primary tumor ulceration and the incidence of elevated serum LDH were significantly higher in our series of mucosal melanomas, compared to the published literature for cutaneous melanoma [2, 12]. A higher proportion of Stage III patients occurred when the primary site was the oral cavity, upper or lower gastrointestinal tract or urological and was relatively lower from the nasal cavity or gynecological sites. Similar results have been published from smaller series [16]. An analysis of patterns of distant metastases demonstrated that lung was the most common site of metastasis (21.0%), followed by liver (18.5%) and distant lymph node (9.0%); these incidences were similar regardless of the anatomic locations of the primary mucosal melanomas. Almost all of the metastases in the liver or lung were multiple, with only 21.5% of lung metastases and 13% of liver metastases presenting as single lesions which might be suitable for surgical excision. There have been many publications about CKIT and BRAF mutation incidence among primary cutaneous melanomas. 872 Lian et al. Volume 28 Issue

6 Annals of Oncology In our series, we observed a 10% incidence of CKIT mutation and 12% incidence of BRAF mutation (Table 1). When examining separate anatomic sites, there was no difference with regard to BRAF mutations (range of 9% to 12.5%) (Table 2). However, there was a wider range of CKIT mutations, ranging from 5.5% for nasopharyngeal melanomas to 17% for urological melanomas (Table 2). The incidence of CKIT mutation incidence from smaller published studies ranged from 4% to 17%, while the incidence of BRAF mutation ranged from 0% to 21% [17 19]. In few small sample size studies, the overall survival of mucosal melanomas from different primary sites was poor, the 5-year survival rates were %, %, %) in nasal pharyngeal and oral, gastrointestinal, gynecological and urological mucosal melanoma, respectively [20]. In our series, we found a similar 1-year, 2- year, 5-year survival rate in those three main primary sites, which indicated they had similar survival pattern despite of different sites, and prognosis factors will be discussed on further analysis. The natural history and patterns of metastases of mucosal melanoma has implications for clinical practice including the primary surgery, systemic adjuvant therapy, and metastatic surveillance. The metastatic behavior of mucosal melanoma showed that metastatic surveillance should include the liver, lung, and regional lymph nodes. The distant metastasis classification is now on further analysis, which could be used as clinical trials stratification criteria and end results reporting. Also, these findings suggest that clinical trials involving mucosal melanomas and the administration of systemic therapy can be applied equally regardless of primary anatomic site. Acknowledgements The authors wish to thank the patients and families for making this study possible. Funding This work was supported by grants from Beijing Municipal Administration of Hospitals Incubating Program (PX ), Beijing Talents Fund ( G186, ZK26, ZK18), National Natural Science Foundation of China ( , ), Beijing Natural Science Foundation ( , , ), the Major State Basic Research Development Program of China (2013CB911004), Beijing Baiqianwan Talents Project, Beijing Municipal Administration of Hospitals Clinical medicine Development of special funding support (ZYLX201603), Beijing Municipal Science & Technology Commission (Z ). Disclosure The authors have declared no conflicts of interest. References Original article 1. Sun CZ, Li QL, Hu ZD et al. Treatment and prognosis in sinonasal mucosal melanoma: a retrospective analysis of 65 patients from a single cancer center. Head Neck 2014; 36: Chi Z, Li S, Sheng X et al. Clinical presentation, histology, and prognoses of malignant melanoma in ethnic Chinese: a study of 522 consecutive cases. BMC Cancer 2011; 11: Dauer EH, Lewis JE, Rohlinger AL et al. Sinonasal melanoma: a clinicopathologic review of 61 cases. Otolaryngol Head Neck Surg 2008; 138: Bishop KD, Olszewski AJ. Epidemiology and survival outcomes of ocular and mucosal melanomas: a population-based analysis. Int J Cancer 2014; 15(134): Lombardi D, Bottazzoli M, Turri-Zanoni M et al. Sinonasal mucosal melanoma: a 12-year experience of 58 cases. Head Neck 2016; 38: E1737 E Spencer KR, Mehnert JM. Mucosal melanoma: epidemiology, biology and treatment. Cancer Treat Res 2016; 167: Chan KK, Chan RC, Ho RS et al. Clinical patterns of melanoma in Asians: 11-year experience in a tertiary referral center. Ann Plast Surg 2016; 77 (Suppl 1): S6 S Edge SB, Compton CC. The American Joint Committee on Cancer: the 7th edition of the AJCC cancer staging manual and the future of TNM. Ann Surg Oncol 2010; 17: Lian B, Si L, Cui C et al. Phase II randomized trial comparing high-dose IFN-alpha2b with temozolomide plus cisplatin as systemic adjuvant therapy for resected mucosal melanoma. Clin Cancer Res 2013; 19: Koivunen P, B ack L, Pukkila M et al. Accuracy of the current TNM classification in predicting survival in patients with sinonasal mucosal melanoma. Laryngoscope 2012; 122: McLaughlin CC, Wu XC, Jemal A et al. Incidence of noncutaneous melanomas in the U.S. Cancer 2005; 103: Chang AE, Karnell LH, Menck HR. The National Cancer Data Base report on cutaneous and noncutaneous melanoma: a summary of 84,836 cases from the past decade. The American College of Surgeons Commission on Cancer and the American Cancer Society. Cancer 1998; 83(8): Balch CM, Wilkerson JA, Murad TM et al. The prognostic significance of ulceration of cutaneous melanoma. Cancer 1980; 45: Breslow A. Thickness, cross-sectional area, and depth of invasion in the prognosis of cutaneous melanoma. Ann Surg 1970; 172: Pollack LA, Li J, Berkowitz Z et al. Melanoma survival in the United States, 1992 to J Am Acad Dermatol 2011; 65: S78 S Grözinger G, Mann S, Mehra T et al. Metastatic patterns and metastatic sites in mucosal melanoma: a retrospective study. Eur Radiol 2016; 26: Zebary A, Jangard M, Omholt K et al. KIT, NRAS and BRAF mutations in sinonasal mucosal melanoma: a study of 56 cases. Br J Cancer 2013; 109: Lyu J, Wu Y, Li C et al. Mutation scanning of BRAF, NRAS, KIT, and GNAQ/GNA11 in oral mucosal melanoma: a study of 57 cases. J Oral Pathol Med 2016; 45: Beadling C, Jacobson-Dunlop E, Hodi FS et al. KIT gene mutations and copy number in melanoma subtypes. Clin Cancer Res 2008; 14: Marija M, Slobodan V, Predrag J et al. Primary mucosal melanomas: a comprehensive review. Int J Clin Exp Pathol 2012; 5: Volume 28 Issue doi: /annonc/mdw