Epidermal autophagy and beclin 1 regulator 1 and loricrin: a paradigm shift in the prognostication and stratification of the American Joint Committee on Cancer stage I melanomas

Abstract

Background: The updated American Joint Committee on Cancer (AJCC) staging criteria for melanoma remain unable to identify high-risk stage I tumour subsets.

Objectives: To determine the utility of epidermal autophagy and beclin 1 regulator 1 (AMBRA1)/loricrin (AMLo) expression as a prognostic biomarker for AJCC stage I cutaneous melanoma.

Methods: Peritumoral AMBRA1 expression was evaluated in a retrospective discovery cohort of 76 AJCC stage I melanomas. AMLo expression was correlated with clinical outcomes up to 12 years in two independent powered, retrospective validation and qualification cohorts comprising 379 AJCC stage I melanomas.

Results: Decreased AMBRA1 expression in the epidermis overlying primary melanomas in a discovery cohort of 76 AJCC stage I tumours was associated with a 7-year disease-free survival (DFS) rate of 81·5% vs. 100% survival with maintained AMBRA1 (P < 0·081). Following an immunohistochemistry protocol for semi-quantitative analysis of AMLo, analysis was undertaken in validation (n = 218) and qualification cohorts (n = 161) of AJCC stage I melanomas. Combined cohort analysis revealed a DFS rate of 98·3% in the AMLo low-risk group (n = 239) vs. 85·4% in the AMLo high-risk cohort (n = 140; P < 0·001). Subcohort multivariate analysis revealed that an AMLo hazard ratio (HR) of 4·04 [95% confidence interval (CI) 1·69-9·66; P = 0·002] is a stronger predictor of DFS than Breslow depth (HR 2·97, 95% CI 0·93-9·56; P = 0·068) in stage IB patients.

Conclusions: Loss of AMLo expression in the epidermis overlying primary AJCC stage I melanomas identifies high-risk tumour subsets independently of Breslow depth. What's already known about this topic? There is an unmet clinical need for biomarkers of early-stage melanoma. Autophagy and beclin 1 regulator 1 (AMBRA1) is a proautophagy regulatory protein with known roles in cell proliferation and differentiation, and is a known tumour suppressor. Loricrin is a marker of epidermal terminal differentiation. What does this study add? AMBRA1 has a functional role in keratinocyte/epidermal proliferation and differentiation. The combined decrease/loss of peritumoral AMBRA1 and loricrin is associated with a significantly increased risk of metastatic spread in American Joint Committee on Cancer (AJCC) stage I tumours vs. melanomas, in which peritumoral AMBRA1 and loricrin are maintained, independently of Breslow depth. What is the translational message? The integration of peritumoral epidermal AMBRA1/loricrin biomarker expression into melanoma care guidelines will facilitate more accurate, personalized risk stratification for patients with AJCC stage I melanomas, thereby facilitating stratification for appropriate follow-up and informing postdiagnostic investigations, including sentinel lymph node biopsy, ultimately resulting in improved disease outcomes and rationalization of healthcare costs.

Figure 1

Study design followed the Cancer Research UK prognostic biomarker roadmap.10 Following initial identification of varied autophagy and beclin 1 regulator 1 (AMBRA1) epidermal expression in formalin‐fixed, paraffin‐embedded American Joint Committee on Cancer (AJCC; 2009) stage I and II primary cutaneous melanoma, an immunohistochemistry (IHC) assay was developed for initial assessment in a retrospective cohort of 76 patients with AJCC stage I melanoma derived from the Newcastle Hospital National Health Service Foundation Trust. Following validation and conversion of the assay to a fully automated IHC system, and the addition of loricrin, two further retrospective cohorts of 218 and 161 patients, respectively, with AJCC (2009) stage I melanomas were analysed for AMBRA1/loricrin expression levels and associated disease‐free survival data. SOP, standard operating procedure.

Figure 2

Relationship between peritumoral autophagy and beclin 1 regulator 1 (AMBRA1) expression and disease‐free survival in the Newcastle discovery American Joint Committee on Cancer (AJCC) stage I melanoma cohort. (a) Representative photomicrographs of immunohistochemical AMBRA1 (pink) or cytokeratin 5 (CK5; brown) staining in normal or matched peritumoral epidermis, where (e) represents the epidermis and (m) identifies melanoma (Breslow depth 1·2 mm, scale bars 100 μm). (b) AMBRA1 levels in the peritumoral epidermis of AJCC stage I melanomas were determined by visual inspection by a pathologist and defined as maintained or decreased. Estimated 7‐year disease‐free survival rates were determined with the Kaplan–Meier method and compared by a two‐sided log‐rank test (81·5% vs. 100%; P < 0·081).

Figure 3

Immunohistochemistry analysis of autophagy and beclin 1 regulator 1 (AMBRA1) protein expression in the James Cook University Hospital American Joint Committee on Cancer stage I cohort. (a) Semi‐quantitative scoring of AMBRA1 (percentage decrease in peritumoral epidermis vs. normal epidermis at the section margins) vs. visual scoring of either maintained or decreased peritumoral AMBRA1 [horizontal bar indicates median with interquartile range (median 11·2% vs. 84·1%; Mann–Whitney, P < 0·001)]. (b) Representative photomicrographs of maintained (low risk) or decreased (high risk) immunohistochemical AMBRA1 staining in peritumoral epidermis at ×100 and ×200 magnification (scale bars 100 μm). (c) Representative photomicrographs of continuous (low risk) or broken (high risk) immunohistochemical loricrin staining in peritumoral epidermis at ×100 and ×200 magnification (scale bars 100 μm). ****indicates value is P < 0.00001.

Figure 4

Relationship between peritumoral autophagy and beclin 1 regulator 1 (AMBRA1)/loricrin expression and disease‐free survival in sentinel lymph node biopsy (SLNB)‐eligible patients. (a) Post‐test probabilities after SLNB for metastatic melanoma in MSLT‐1 (Multicenter Selective Lymphadenectomy Trial) intermediate‐thickness cohort. (b) Post‐test probabilities after analysis of AMBRA1/loricrin expression in SLNB‐eligible American Joint Committee on Cancer stage I melanoma samples.

Figure 5

Relationship between peritumoral autophagy and beclin 1 regulator 1 (AMBRA1)/loricrin expression and disease‐free survival in the validation cohorts. AMBRA1 and loricrin levels in the peritumoral epidermis of American Joint Committee on Cancer stage I melanomas were determined by visual inspection by a pathologist and defined as maintained or decreased. Ten‐year disease‐free survival rates were determined with the Kaplan–Meier method and compared by a two‐sided log‐rank test in both the (a) James Cook University Hospital [JCUH; 84·8% vs. 98·1% (P < 0·001)] and (c) University Hospital of North Durham [UHND; 88·8% vs 97·5% (P = 0·033)] cohorts. Assay performance and univariate and multivariate cox regression analysis of disease‐free survival in the (b) JCUH and (d) UHND cohorts. CI, confidence interval.

Figure 6

Relationship between peritumoral autophagy and beclin 1 regulator 1 (AMBRA1)/loricrin expression and disease‐free survival in the combined validation cohorts. AMBRA1 and loricrin levels in the peritumoral epidermis of American Joint Committee on Cancer (AJCC) stage I melanomas were determined by visual inspection by a pathologist and defined as maintained or decreased. Twelve‐year disease‐free survival (DFS) rates were determined with the Kaplan–Meier method and compared by a two‐sided log‐rank test in the combined validation cohort (86·1% vs. 97·9%; P < 0·001). (b) Assay performance and univariate and multivariate Cox regression analysis of DFS in the combined cohort. (c) Twelve‐year DFS rates were determined with the Kaplan–Meier method and compared by a two‐sided log‐rank test in AJCC stage IB melanomas of the combined validation cohort (80·5% vs. 96·2%; P < 0·001). (d) Assay performance and univariate and multivariate Cox regression analysis of DFS in AJCC stage IB melanomas of the combined cohort. CI, confidence interval.