Why Your New Cancer Biomarker May Never Work: Recurrent Patterns and Remarkable Diversity in Biomarker Failures
Author's Affiliation: The Sidney Kimmel Comprehensive, Cancer Center at Johns Hopkins, Baltimore, Maryland.Cancer Research (Impact Factor: 9.33). 11/2012; 72(23). DOI: 10.1158/0008-5472.CAN-12-3232
Less than 1% of published cancer biomarkers actually enter clinical practice. Although best practices for biomarker development are published, optimistic investigators may not appreciate the statistical near-certainty and diverse modes by which the other 99% (likely including your favorite new marker) do indeed fail. Here, patterns of failure were abstracted for classification from publications and an online database detailing marker failures. Failure patterns formed a hierarchical logical structure, or outline, of an emerging, deeply complex, and arguably fascinating science of biomarker failure. A new cancer biomarker under development is likely to have already encountered one or more of the following fatal features encountered by prior markers: lack of clinical significance, hidden structure in the source data, a technically inadequate assay, inappropriate statistical methods, unmanageable domination of the data by normal variation, implausibility, deficiencies in the studied population or in the investigator system, and its disproof or abandonment for cause by others. A greater recognition of the science of biomarker failure and its near-complete ubiquity is constructive and celebrates a seemingly perpetual richness of biologic, technical, and philosophical complexity, the full appreciation of which could improve the management of scarce research resources. Cancer Res; 72(23); 1-5. ©2012 AACR.
[Show abstract] [Hide abstract]
- "Along with the recent trend on computational methods for evaluating biomarkers, we here raise new and rather underdeveloping measures on two cases: (1) two or more studies for a single biomarker, and (2) one study for multiple biomarkers. Now various studies for a single biomarker can be gener- ated [8,15]. For example, even by the same research group, studies can be repeated at a certain time interval. "
ABSTRACT: Biomarkers are vital to detect diseases in various clinical stages. A variety of cancer serum biomarkers are already known, while for more accurate cancer-type detection, there required more rigorous evaluation manners, especially computational evaluation measures, for biomarkers. In this review, we first show three typical pitfalls in finding biomarkers and their examples, after briefly presenting standard five clinical biomarker screening phases by National Cancer Institute. We then introduce current computational biomarker evaluation measures, including current, standard methods with their intrinsic features. We further show an up-to-date list of existing cancer serum biomarkers, pointing out several issues, being caused by the limitations of current biomarker evaluation approaches. Finally we discuss the current attempts to develop new, statistically robust, computational serum-based biomarker measures in terms of specificity to each of various cancer types.
[Show abstract] [Hide abstract]
- "The resulting multi-gene prognostic biomarkers (sometimes called signatures) can identify patient subgroups that would be particularly likely to derive benefit from more intense therapy [5,6]. However, there have been numerous challenges in the development of clinically-useful biomarkers; most published biomarkers fail to enter routine clinical practice . "
ABSTRACT: Background The reproducibility of transcriptomic biomarkers across datasets remains poor, limiting clinical application. We and others have suggested that this is in-part caused by differential error-structure between datasets, and their incomplete removal by pre-processing algorithms. Methods To test this hypothesis, we systematically assessed the effects of pre-processing on biomarker classification using 24 different pre-processing methods and 15 distinct signatures of tumour hypoxia in 10 datasets (2,143 patients). Results We confirm strong pre-processing effects for all datasets and signatures, and find that these differ between microarray versions. Importantly, exploiting different pre-processing techniques in an ensemble technique improved classification for a majority of signatures. Conclusions Assessing biomarkers using an ensemble of pre-processing techniques shows clear value across multiple diseases, datasets and biomarkers. Importantly, ensemble classification improves biomarkers with initially good results but does not result in spuriously improved performance for poor biomarkers. While further research is required, this approach has the potential to become a standard for transcriptomic biomarkers.
[Show abstract] [Hide abstract]
- "A miRNA signature trial has been recently announced (MRX34 trial). A key issue for miRNA signatures for cancer is their validation in diverse cohorts (Kern, 2012). Such task generates difficulties even though many public cohorts are available. "
ABSTRACT: MicroRNAs (miRNAs) play a key role in post-transcriptional regulation of RNAm levels. Their function in cancer has been studied by high throughput methods generating valuable sources of public information. Thus miRNA signatures predicting cancer clinical outcomes are emerging. An important step to propose miRNA-based biomarkers before clinical validation is their evaluation in independent cohorts. Although it can be carried out using public data, such task is time consuming, and requires a specialized analysis. Therefore to aid and simplify the evaluation of prognostic miRNA signatures in cancer, we developed SurvMicro, a freely and easy to use web tool that assesses miRNA signatures from publicly available miRNA profiles using multivariate survival analysis. SurvMicro is composed of a wide and updated database of more than 40 cohorts in different tissues and a web tool where survival analysis can be done in minutes. We presented evaluations to portrait the straightforward functionality of SurvMicro in Liver and Lung cancer. To our knowledge SurvMicro is the only bioinformatic tool that aids the evaluation of multivariate prognostic miRNA signatures in cancer. SurvMicro and its tutorial are freely available at http://bioinformatica.mty.itesm.mx/SurvMicro.
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.