Sites of distant metastasis in Merkel cell carcinoma differ by primary tumor site and are of prognostic significance: A population-based study in the Surveillance, Epidemiology, and End Results database from 2010 to 2016

To the Editor: Merkel cell carcinoma (MCC) is highly aggressive, with a propensity for recurrence and distant metastasis. Sentinel lymph node biopsy critical workup; however, optimal use of imaging unclear.1Schmults C.D. Blitzblau R. Aasi S.Z. et al.Merkel Cell Carcinoma, Version 1.2020, NCCN Clinical Practice Guidelines in Oncology. National Comprehensive Cancer Network.J Natl Compr Canc Netw. 2018; 16: 742-774PubMed Google Scholar A large institutional registry recently identified that liver metastases were more common head/neck primary compared other sites, which could guide workup surveillance strategies.2Lewis C.W. Qazi J. Hippe D.S. al.Patterns 215 patients: implications prognosis surveillance.Cancer Med. 2020; 9: 1374-1382Crossref PubMed Scopus (22) Thus, we sought to investigate effect tumor site on metastasis patterns national cohort, hypothesizing relationship would exist. We patients MCC (code 8247/3) Surveillance, Epidemiology, End Results (SEER)-18 registries. Our inclusion criteria consisted metastatic disease upon initial presentation based American Joint Committee 8th Edition staging, diagnosed between 2010 2016. cohort was stratified by lesion site, trunk, head/neck, upper/lower extremities, sites encompassing visceral or nodal disease. used χ2 tests Student t (2-tailed P values) assess differences clinical pathologic characteristics distributions performed Kaplan-Meier analysis analyze overall survival disease-specific (DSS). multivariate Cox proportional hazards independent prognosticators DSS. 331 M1 predominantly male White (Table I). Patients had higher proportion (42.3%, = .0003) relative trunk bone (36.9%, .0049). Overall 5-year 11.2%, DSS 16.2%. Increasing age brain poorer II).Table ICohort siteVariable∗Data are presented as number (%) unless indicated otherwise.Head/neckTrunkUpper/lower extremitiesOther†Owing nature coding database, “other” category contains overlapping skin those an unknown where first discovered location.P valueNumber (% cohort)97 (29.3)38 (11.5)91 (27.5)105 (31.7).0039Age at diagnosis, mean (SD), y78.5 (9.7)71.7 (11.2)77.3 (11.6)70.8 (13.5)Sex.003 Male82 (84.5)26 (68.4)56 (61.5)77 (73.3) Female15 (15.5)12 (31.6)35 (38.5)28 (26.7)Race.107 White95 (98.0)33 (86.8)85 (93.4)98 (93.3) African American1 (1.0)2 (5.3)5 (5.5)1 (1.0) Other/unknown1 (1.0)3 (7.9)1 (1.1)6 (4.7)Tumor size<.0001 0-10 mm9 (9.3)1 (2.6)4 (4.4)35 (33.3) 11-20 mm14 (14.4)011 (12.1)1 21-30 mm16 (16.5)7 (18.4)8 (8.8)2 (1.9) 31-40 mm6 (6.2)5 (13.2)9 (9.9)0 41-50 mm3 (3.1)2 (5.3)8 >50 (3.1)10 (26.3)17 (18.7)1 Unknown46 (47.4)13 (34.2)34 (37.3)64 (60.9)Bone metastasis.0049 Yes24 (24.7)14 (36.9)15 (16.5)15 (14.3) No70 (72.2)23 (60.5)73 (80.2)77 Unknown3 (3.1)1 (2.6)3 (3.3)13 (12.4)Brain metastasis.0315 Yes3 (2.6)02 No89 (91.7)36 (94.8)88 (96.7)89 (84.8) Unknown5 (5.2)1 (3.3)14 (13.3)Liver metastasis.0003 Yes41 (42.3)5 (13.2)21 (23.1)22 (20.9) No52 (53.6)32 (84.2)67 (73.6)70 (66.7) Unknown4 (4.1)1 (12.4)Lung metastasis.272 Yes13 (13.4)7 (18.4)18 (19.8)13 (12.4) No79 (81.4)30 (79.0)70 (76.9)81 (77.1) (3.3)11 (10.5)∗ Data otherwise.† Owing location. Open table new tab Table IICox (DSS) carcinomaVariableDSS hazard ratio (95% CI)P valueAge diagnosis1.03 (1.02-1.05)<.0001Sex Female (Ref)1.0 Male1.14 (0.83-1.56).417Race American1.82 (0.78-3.73).155Primary Head neck Trunk1.39 (0.86-2.27).181 Upper lower extremities Other1.05 (0.72-1.53).805 Other∗Owing location.1.53 (1.01-2.32).044Tumor size mm mm1.48 (0.79-2.78).222 mm1.07 (0.60-1.92).818 mm0.84 (0.42-1.68).627 mm1.63 (0.82-3.25).166 mm1.61 (0.86-3.02).136Bone No Yes1.18 (0.84-1.67).083Brain Yes3.85 (1.58-9.38).0030Liver Yes1.86 (1.37-2.52)<.0001Lung Yes1.12 (0.77-1.64).555CI, Confidence interval; Ref, reference.∗ CI, reference. corroborates findings Lewis al,2Lewis also suggesting MCC, associated increased metastasis, generalizing this finding cohort. Additionally, demonstrate our (although rare), but not lung bone, These may further research regarding MCC. Current guidelines suggest whole-body positron emission tomography (PET)/computed (CT) PET/magnetic resonance (MRI), CT chest, abdomen, pelvis contrast, without MRI, evaluation regional disease, when clinically indicated.1Schmults however do all node-negative should be screened PET/CT because likely change staging management features potential advanced (large size, lymphovascular invasion, immunosuppression). Considering findings, unexplained function abnormalities particularly raise suspicion decisions. Furthermore, methods differ sensitivity specificity detection, limited data greater utility detection studied malignancies) CT.3Hawryluk E.B. O'Regan K.N. Sheehy N. al.Positron tomography/computed carcinoma: study 270 scans 97 Dana-Farber/Brigham Women's Center.J Am Acad Dermatol. 2013; 68: 592-599Abstract Full Text PDF (66) In addition, area interest PET/MRI, has shown better PET/CT, although these MCC.4Hong S.B. Choi S.H. Kim K.W. al.Diagnostic performance [(18)F]FDG-PET/MRI malignancy: systematic review meta-analysis.Eur Radiol. 2019; 29: 3553-3563Crossref (13) This relevant primary, cost-effectiveness important consider. given rarity studies focus selection MRI. Ultimately, future prospective will needed clarify strategies consideration can research. Limitations include retrospective lack details immune status location beyond liver, lung, brain. from largely predate checkpoint inhibitors optimizing early increasingly burden melanoma been influence treatment response progression-free blockade, case MCC.5Davis E.J. Perez M.C. Ayoubi al.Clinical correlates anti-PD-1–based therapy melanoma.J Immunother. 42: 221-227Crossref (16) provide insight into help studies.