Table 1. Characteristics of Included Studies.
| Study | Index prognostic factor | Design, setting, and population | Sampling technique (dates of melanoma diagnosis) | Participants by AJCC stage, No. (%) | Male, No. (%) | Age, median, y (follow-up duration, median, y) | Breslow thickness, median, mm | Ulceration, No. (%) | Anatomical sites, No. (%) | Follow-up and imaging protocol | Outcomes blinded to GEP result |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Hsueh et al,25 2017 (US) | DecisionDx-Melanoma (Castle Biosciences Inc [which also provided funding]) | Observational, prospective registry, multicenter (n = 11); hospital and community based; patients with resected melanoma with a successful GEP resulta | Nonprobability, convenience (NR) | 209 (66) Stage I; 73 (23) stage II; 36 (11) stage III | 176 (55) | 58 (1.5 Without event and 1 with event) | 1.2 | Present: 58 (18); unknown: 26 (8) | Head/neck: 58 (18); trunk: 86 (27); extremity: 178 (55) | Nob | Nob |
| Greenhaw et al,29 2018 (US) | DecisionDx-Melanoma (Castle Biosciences Inc; funded by Zitelli and Brodland, PCc) | Observational, mixed retrospective and prospective cohorts; single center, community based; patients with resected melanoma with previous GEP testing performed as part of clinical care and patients with melanoma with known metastatic disease who subsequently underwent GEP testing | Nonprobability, convenience (NR) | 219 (86) Stage I; 37 (14) stage II | 160 (63) | 68 For class 1 and 72 for class 2 (1.9 [mean]) | 0.4 (Class 1); 2 (class 2) | Present: 26 (10); unknown: 2 (1) | NR | Retrospective cohort: non–rule based; prospective cohort: rule based by GEP test scorec | Noc |
| Zager et al,27 2018 (US) | DecisionDx-Melanoma (Castle Biosciences Inc [which also provided funding]) | Observational, retrospective cohort of archival specimens; multicenter (n = 16), hospital and community based; patients with resected stage I-III melanoma with minimum of 5 y of follow-up if no recurrence | Nonprobability, convenience (2000-2014) | 264 (50) Stage I; 93 (18) stage II; 166 (32) stage III | NR | 59 (7.5 Without event; 1.2 with event) | 1.2; Unknown in 4 (1%) | Present: 133 (26); unknown: 81 (15) | NR | No | Yes |
| Keller et al,26 2019 (US) | DecisionDx-Melanoma (Castle Biosciences Inc, funded by St. Louis University Cancer Center) | Observational, prospective cohort; single center, hospital based; patients with melanoma undergoing WLE and SLN biopsya | Nonprobability, convenience (2013-2015) | 96 (60) Stage I; 40 (25) stage II; 23 (15) stage III | 98 (62) | 59 (3.5) | 1.4 | Present: 38 (24) | Head/neck: 26 (16); trunk: 68 (43); extremity: 65 (41) | Rule based by stage of disease | No |
| Podlipnik et al,28 2019 (Spain) | DecisionDx-Melanoma (Castle Biosciences Inc [which also provided funding]) | Observational, prospective cohort; multicenter (n = 5), hospital based; patients with resected stage IB-II melanoma (85% underwent SLN biopsy)d | Nonprobability, convenience (2015-2016)d | 62 (72) Stage IB or IIA; 24 (28) stage IIB or C | 40 (47) | 59 (2.2) | 2.5 (Mean) | Present: 26 (30) | Head/neck: 11 (13); trunk: 37 (43); legs: 21 (24); arms: 12 (14); acral: 5 (6) | Nod | Nod |
| Koelblinger et al,31 2018 (Germany) | MelaGenix (NeraCare GmbH [funding not reported]) | Observational, case-control (retrospective); hospital based; patients with melanoma ≤1 mm with recurrence (cases) or disease free (controls) | NR (NR) | 88 (100) Stage I | NR | NR (3.7) | 0.8 (Cases); 0.6 (controls) | NR | NR | NR | Yes |
| Amaral et al,30 2020 (Germany) | MelaGenix (NeraCare GmbH [which also provided funding along with Eberhard-Karls University of Tuebingen])e | Observational, retrospective analysis of archival registry; single center, hospital based; patients with resected stage II melanoma with available tissue (80% underwent SLN biopsy) | Nonprobability, consecutive (2000-2016)e | 245 (100) Stage II | 134 (55) | 70 (3.4) | 3.0 | Present: 142 (58) | NR | Rule based by stage of diseasee | Yese |
Abbreviations: AJCC, American Joint Committee on Cancer; GEP, gene expression profile; NR, not reported; SLN, sentinel lymph node; WLE, wide local excision.
a
There were partially overlapping patient populations (ie, a proportion of patients described by Keller et al26 are included in the study by Hsueh et al25).
b
Date of email communication with study author was December 11, 2019.
c
Date of email communication with study author was January 21, 2020.
d
Dates of email communication with study author were December 13, 2019; and January 22, 2020; and February 4, 2020.
e
Date of email communication with study author was January 22, 2020.