Not All Melanomas are Created Equal: A Review and Call for More Research Into Nodular Melanoma

C. Dessinioti; A.C. Geller; D.C. Whiteman; C. Garbe; J.J. Grob; J.W. Kelly; R.A. Scolyer; R.V. Rawson; A. Lallas; G. Pellacani; A.J. Stratigos

Disclosures

The British Journal of Dermatology. 2021;185(4):700-710.

Abstract and Introduction

Abstract

Among the histogenic subtypes of melanoma, nodular melanoma (NM) is the major contributor for thicker and fatal melanomas and it has been associated with melanoma-specific death in thin tumours, highlighting an important subgroup of 'aggressive thin' melanomas. This review provides a synthesis of the distinct characteristics of NM, with respect to epidemiology and risk factors, clinical presentation, histopathology, molecular and dermoscopic aspects, and screening practices. The real challenges are to find better biomarkers of aggressiveness and to know whether the control of such aggressive melanomas can be influenced by targeted interventions such as early detection, drug interventions and preventive strategies.

Introduction

Cutaneous melanoma (CM) accounts for approximately 10% of skin cancers in light-skinned populations, but causes the most skin cancer-related deaths.^[1] Melanoma is a heterogenous tumour comprised of biologically distinct subtypes based on their cell of origin, role of ultraviolet radiation exposure, pattern of somatic mutations and precursor lesion.^[2] The four major subtypes of invasive melanoma distinguished by clinical and histopathological features are: superficial spreading melanoma (SSM) (approximately 41–56%), followed by nodular melanoma (NM) (16%^[3] to 25·6%^[4]), lentigo maligna melanoma (LMM) (2·7–14%) and acral lentiginous melanoma (ALM) (1–5% in non-Hispanic white populations and higher proportions but similar absolute incidence in Asian and African American populations).^[5] When 'melanoma not otherwise specified' is included in the reported subtypes, the rates of NM range from 7% to 16·5%.^[3,6–12]

The NM subtype is the main contributor for thicker and fatal melanomas,^[9,11] thus a persistent burden on public health.^[9,12,13] The median Breslow thickness of NM has increased over time, while it decreased for SSM, as reported in a Surveillance, Epidemiology and End Results (SEER)-9 registry study (1989–2009).^[9] The greater Breslow thickness associated with NM may be attributed in part to obstacles that reduce the likelihood of early detection as well as inherent biological characteristics, such as a faster growth rate,^[14,15] and points to distinct clinical and biological characteristics from its outset compared with other melanoma subtypes.

This review presents a comprehensive synthesis of the current knowledge of characteristics of NM that underlie its distinctive aggressive nature. In this context, the evidence for NM available in the literature was reviewed, with regard to epidemiology and survival, risk factors, clinical presentation, histopathology, molecular differences, characteristics with dermoscopy and reflectance confocal microscopy, and screening practices.

1 2 3 4 5 6 7 8 9

Next Section

Table 1. Studies reporting survival outcomes in association with the nodular melanoma subtype
Publication	Type of study, country, period	n NM (N total)	CM types included	Follow-up time	OS of NM	MSS of NM	Other survival outcomes of NM
El Sharouni, 2020³¹	Retrospective, Dutch Nationwide Pathology Registry, 2000–2014	7095 (48 361)	NM, SSM, LMM, ALM Stage I–III	Median: 73·8 months	OS: Association with NM for T1 cases without ulceration, HR 1·96 (95% CI 1·58–2·45), and for T1 cases with ulceration HR 3·46 (2·17–5·50), adjusted for age, sex, localization, stage No association with NM in thicker melanomas	NR	NR
Claeson, 2020²⁹	Population-based Queensland Cancer Registry, Australia, 1995–2014	39 (1613)	Thin (≤ 1·00 mm) localized	NR	NR	No association with NM: OR 1·03 (95% CI 0·46–2·30), adjusted for thickness, location, ulceration, level of invasion	NR
Dessinioti, 2019³⁰	Retrospective, multicentre, Europe, Australia, USA, 2006–2015	5062 (20 132)	NM, SSM	Median: 32·1 months	NR	Association with NM for T1 cases, HR 2·20 (95% CI 1·28–3·78), adjusted for age, sex, thickness, ulceration, centre No association with NM in localized or in thicker melanomas	NR
Lattanzi, 2019²²	Retrospective, registry and hospital-based, SEER, NYU, USA	SEER: 21 399 (118 508) NYU: 510 (1621)	NM, SSM SEER: stage I–III	Median: SEER: 68 months NYU: 73 months	NR	Association with NM, SEER, HR 1·55 (95% CI 1·41–1·70), adjusted for age, sex, year of diagnosis, thickness, ulceration, stage NYU, HR 1·47 95% CI 1·05–2·07), adjusted for age, sex, year of diagnosis, thickness, ulceration, mitotic index, stage	DFS: association with NM, NYU, HR 1·45 (95% CI 1·11–1·89)
Fujisawa, 2019¹⁰	Retrospective, survey of the Japanese Skin Cancer Society, Japan, 2005–2017	458 (4594)	In situ, invasive	NR	NR	No association with NM: HR 1·38 (95% CI 0·95–1·99), adjusted for age, sex, T status, ulceration, regression, vitiligo, satellite, N status, distant metastasis	NR
Robsahm, 2018⁴	Retrospective, Norwegian Malignant Melanoma Registry, 2008–2012	1527 (8087)	Invasive	Median: 4 years (range 0·08–7·5)	NR	No association with NM (ref SSM): HR 1·01 (95% CI 0·79–1·29), adjusted for age, sex, T stage, location, ulceration, clinical stage, second primary	NM accounted for the dominant proportion of fatal melanomas (55·3% in women, 64·6% in men)
Green, 2018²³	Retrospective, hospital-based, England, Australia, 2010–2014	347 (1488)	Stage IB/II	For Australia, 2·9–6·9 years	NR	NR	For thinner ≤ 2 mm: DFS 82% for NM, DFS 91% SSM (P = 0·051) For thicker > 2 mm: no difference in DFS
Carrera, 2017²⁵	Retrospective, multicentre, South Europe, 1985–2010	20 (269)	Acral melanoma In situ, invasive	Mean: 53·59 months (± 44·06)	NR	Mean MSS 54·9 months	Mean DFS 49·3 months
Hugdahl, 2016²⁶	Prospective, hospital-based, Norway, 1998–2008	248 (248)	NM	Median: 31 months (range 0–131)	NR	Independent predictors of worse survival: thickness, ulceration, mitotic count, BRAF-V600E expression	Worse MSS for those positive for BRAF-V600E expression
Mar, 2013¹¹	Retrospective, Victorian Cancer Registry, 1989, 1994, 1999, 2004	782 (5775)	Invasive	1989–2010	NR	Association with NM (vs. SSM). HR 1·5 (95% CI 1·2–1·8), adjusted for age, sex, thickness	NM was the main contributor subtype to melanoma-specific deaths (43%)
Green, 2012²⁸	Retrospective, Queensland Cancer Registry, 1982–2006	458 (26 736)	Thin ≤ 1 mm	1–20 years	NR	20–year MSS 84·9% Association with NM. HR 1·80 (95% CI 1·3–2·6), adjusted for year, age, sex, thickness, location, Clark level	NR
Egger, 2012²⁷	Retrospective review of a randomized prospective clinical trial, North America, 1997–2003	736 (736)	NM ≥ 1 mm with clinically negative lymph nodes	Median: 67 months	5-year OS 86·4% without ulceration/65·5% with ulceration Independent prognostic factors: thickness, ulceration, SLN status	NR	5-year DFS 83·9% without ulceration/67·9% with ulceration Independent prognostic factors: BT, ulceration, SLN status
Shaikh, 2012³	Retrospective population-based cohort study, SEER-9, 1978–2007	8973 (111 478)	Invasive	Minimum follow-up: 10 years	NR	NR	NM 37% of fatal cases SSM 46% of fatal cases
Criscione, 2010⁷	Retrospective, population-based registries, SEER-17, 1988–2006	7% NM of 153 124	In situ, invasive	NR	NR	NR	NM and SSM made up a similar proportion of fatal cases (18% vs. 17%)

ALM, acral lentiginous melanoma; BT, Breslow thickness; CI, confidence interval; CM, cutaneous melanoma; DFS, disease-free survival; HR, hazard ratio; LMM, lentigo maligna melanoma; MSS, melanoma-specific survival; NM, nodular melanoma; NR, not reported; NYU, New York University; OR, odds ratio; OS, overall survival; SEER, Surveillance, Epidemiology and End Results; SLN, sentinel lymph node; SSM, superficial spreading melanoma; T1, with Breslow thickness ≤ 1 mm