医学
危险系数
数字聚合酶链反应
肿瘤科
荟萃分析
子群分析
科克伦图书馆
内科学
生物标志物
梅德林
系统回顾
置信区间
聚合酶链反应
生物
生物化学
基因
作者
Lara Gracie,Yi Pan,Eshetu G. Atenafu,Douglas G. Ward,M.S.L. Teng,Lallit Pallan,Neil M. Stevens,Leila Khoja
标识
DOI:10.1016/j.ejca.2021.09.019
摘要
Circulating tumour DNA (ctDNA) is an emerging biomarker in melanoma. We performed a systematic review and meta-analysis to explore its clinical utility as a prognostic, pharmacodynamic (PD) and predictive biomarker.A systematic search was conducted from Jan 2015 to April 2021, of the electronic databases PubMed, Cochrane Library and Ovid MEDLINE to identify studies. Studies were restricted to those published in English within the last 5 years, evaluating ctDNA in humans in ≥10 patients. Survival data were extracted for meta-analysis using pooled treatment effect (TE), i.e. log hazard ratios (HRs) and corresponding standard error of TE for progression-free survival or overall survival differences in patients who were ctDNA positive or negative. PRISMA statement guidelines were followed.A meta-analysis of 19 studies grouped according to methodology of ctDNA detection, revealed a combined estimate for HR of progression-free survival (13 studies using droplet digital Polymerase Chain Reaction (ddPCR) methodology (N = 1002) of 2.10 (95% CI: 1.71-2.59) revealing a poorer prognosis when ctDNA was detected. This result was confirmed in the smaller analysis of (non-ddPCR, N = 347) five studies: HR = 2.45 (95% CI: 1.29-4.63). Similar findings were found in the overall survival analysis of nine studies (ddPCR methodology, N = 841) where the combined HR was 2.78 (95% CI: 2.21-3.49) and of the five studies (non-ddPCR methodology, N = 326) where the combined HR was 2.58 (95% CI: 1.74-3.84). Serial ctDNA levels on treatment showed a pharmacodynamic role reflecting response or resistance earlier than radiological assessment.Circulating tumour DNA is a predictive, prognostic and PD biomarker in melanoma. Technical standardisation of assays is required before clinical adoption.
科研通智能强力驱动
Strongly Powered by AbleSci AI