Cancer Biomarkers and Molecular Theranostics

Full Article

ESK Ma, CLP Wong

Hong Kong J Radiol 2010;13(Suppl):S42-50

Cancer biomarkers have evolved from assays based on proteins, hormones and enzymes to molecular assays based on DNA or RNA. These molecular cancer biomarkers have broad clinical applications in disease screening, diagnosis, classification, offering a prognosis, risk stratification, treatment selection and monitoring. The prevailing trend in oncology is to harness molecular biomarkers to predict treatment efficacy or toxicity in the individual patient, and thereby guide the choice of treatment. The term theranostics was recently coined to indicate this marriage between an individual’s specific diagnosis and tailored therapy. Before therapeutics can be directly linked to diagnostics, the analytical validity, clinical validity, and clinical utility of the test in question should be carefully considered, not to mention the consequential ethical and financial implications. Recourse to predictive cancer markers with a view to targeted molecular therapy may have positive and negative connotations. The best known examples of positive predictors are the EGFR gene mutation in lung cancer and the presence of HER2 gene amplification in breast cancer, whilst the KRAS gene mutation is a negative predictor in metastatic colorectal cancer. Development of companion diagnostics is now double fuelled by the ever-expanding repertoire of agents used for targeting; detection of EML4-ALK gene fusion for consideration of ALK inhibitor therapy in lung cancer being a recent example of this phenomenon. Apart from such predictive markers, pharmacogenomics is another important facet of cancer theranostics. For instance, thymidylate synthase expression or genotype is related to the response to 5-fluorouracil and related compounds. It is envisaged that in the future, emerging diagnostic tools — including the next generation of sequencing technology and array-based comparative genomic hybridisation applied to the cancer genome — will further advance personalised oncology.





腫瘤生物標記物的發展已經從蛋白質、賀爾蒙及酶演變至DNA或RNA的分子測定法。這些分子腫瘤 生物標記物應用廣泛,可用作為對疾病的篩選、診斷、分類、預後評估、風險評級、選擇治療方法 及監察。現今腫瘤學的趨勢是利用分子生物標記物來預測個別病人的治療效果或所產生的毒性,從 而引導為病人選擇治療計劃。「治療診斷學」這詞正是用來表達將個人病情的診斷以及度身訂造的 治療計劃合併的含義。在治療連繫診斷前,必需小心考慮測試本身的分析可信性、臨床有效性及應 用;此外亦要考慮引起的道德及費用負擔問題。預測性腫瘤標記物在標靶分子治療中可分為陽性及 陰性兩種。陽性預測因子的最佳例子如肺癌中的EGFR基因突變,或乳癌中的HER2基因擴增,這些 基因都為標靶治療起了陽性預測作用。另一方面,KRAS基因突變在轉移性的結直腸癌中起了一個陰性預測作用。不斷擴大的標靶治療藥物庫加倍催谷了配套檢測的發展,最新的例子是利用EML4-ALK 融合基因的檢定作治療肺癌中ALK抑制劑治療的考慮。除了這些預測性標記物外,藥物基因組學是 腫瘤治療診斷學的另一重要範疇,其中一個例子是胸腺嘧啶合成酶的表達水平或基因型與5-氟尿嘧 啶及其合成物的反應有關。可以想像得到,新興的診斷工具(包括新一代的基因測序技術及應用在 癌症基因組的基因芯片的比較基因組雜交技術)會進一步提升個人化腫瘤治療。