AVAST-M melanoma cohort

This study made use of individual patient-level and transcriptomic data from the phase III adjuvant AVAST-M study, investigating the role of the angiogenesis inhibitor bevacizumab vs placebo in high-risk primary CM^18,19. One thousand three hundred forty-three stage American Joint Committee on Cancer stage IIB (T3bN0M0 and T4aN0M0), IIC (T4bN0M0) and III (TxN1–3M0) cutaneous melanoma (seventh edition AJCC²⁰) were recruited to the study over the period July 18, 2007–March 29, 2012, as previously described. The study (including the collection of DNA and RNA) was ethically approved in accordance with the Declaration of Helsinki (REC reference number 07/Q1606/15, 16th March 2007). Participants provided written informed consent to sampling of their tumour blocks during study recruitment (and prior to the investigational systemic therapy).

All study participants underwent a sentinel LN node biopsy, and if positive proceeded to a completion LN clearance as per the study protocol. Demographic (including gender, age, centre, as well as pathologic data (site of primary, Breslow depth, ulceration, LN involvement and BRAF/NRAS mutation by pyrosequencing)) was collected at the time of randomisation. Data were also collected on the timing, presence/absence and site/s of distant metastases (according to the findings from CT scanning). Data on overall and progression-free survival were collected with a minimum of 6 years follow up.

RNA-sequencing data was available on 204 primary melanoma samples of which 10 samples were removed from the downstream analyses owing to lack of data on all the clinical covariates, and 175 regional LN samples of which 32 samples removed from the downstream analyses owing to a lack of data on all clinical covariates (Supplementary Fig. S1).