DP02: AMBLor: an immunohistochemical prognostic biomarker to identify low-risk nonulcerated cutaneous AJCC stage I/II melanomas

Philip Sloan,^1,2,3 Tom Ewen,¹ Gyorgy Paragh,⁴ Paul Bogner,⁴ Akhtar Hussain,³ Niki Stefanos,⁵ Paul Barrett,⁶ Sonia Mailer,⁷ Grant McArthur,⁷ Jane Meaney,² Richard Tuson² and Penny Lovat^1,2

¹Translation and Clinical Research Institute, Newcastle University, Newcastle upon Tyne, UK; ²AMLo Biosciences Ltd, Newcastle upon Tyne, UK; ³Pathology, Royal Victoria Infirmary, Newcastle upon Tyne, UK; ⁴Dermatology, Roswell Park Comprehensive Cancer Centre, Buffalo, NY, USA; ⁵Pathology, Addenbrookes Hospital, Cambridge, UK; ⁶Pathology, University Hospitals of North Durham, Durham, UK and ⁷Pater McCallum Cancer Centre, Melbourne, Australia

With a predicted rise in incidence of +60% by 2040, melanoma is an increasing world health problem. Furthermore, the 8th edition of American Joint Committee on Cancer (AJCC) staging criteria are unable to identify subsets of patients with stage I/II melanomas with low risk of disease progression, emphasizing the acute need for credible prognostic biomarkers to stratify patient follow-up based on personalized risk and to aid increasing pressure on healthcare management and resources. The combined immunohistochemical (IHC) expression of AMBRA1 and loricrin in the epidermis overlying nonulcerated AJCC stage I melanomas has been identified as a robust prognostic biomarker and valuable pre-sentinel lymph node biopsy (SLNB). In the present multicentre validation study, retrospective analysis of AMBRA1 and loricrin using a prevalidated IHC kit (AMBLor), developed by AMLo Biosciences Ltd was performed in a mixed cohort of 411 AJCC stage I and II nonulcerated AJCC stage II cutaneous melanomas. Clinical follow-up data ranged from 60 to 287 months and each cohort was powered to represent rates of metastasis of 10% for AJCC stage I or up to 20% for stage II disease. Clinical utility analysis revealed retention of AMBLor (AMBLor low risk) was associated with significantly increased disease-free survival of 97% vs. 87% for patients with melanomas in which AMBLor was lost (P = 0·01; hazard ratio 0·20, 95% confidence interval 0·09–0·42), and with a negative predictive value of 97·1%. A budget impact model was subsequently constructed using parameters from National Health Service reference costs and a freedom of information request, involving 75 trusts and published papers. The model assumed patients with AJCC stage IA AMBLor low-risk melanoma are stepped-down from the NG14 National Institute for Health and Care Excellence melanoma guidelines follow-up regimen of four in 1 year to one at the end of the first year and with AJCC stage IB–IIB AMBLor low-risk melanomas reduced from 16 visits over 5 years to four annual reviews. Attendant reductions in computed tomography scans, SLNB and complications were modelled with pathology costs demonstrating an average saving of £175·18 per patient. Collectively, these data suggest AMBLor as a marker to identify a genuinely low-risk subsets of patients with AJCC stage I/II melanomas. Furthermore, use of AMBLor may serve as a means of stratifying patients for reduced follow-up and/or SLNB, with the potential for significant savings on healthcare resources.