Margaret Pepe et al. have published a commentary in the Journal of the National Cancer Institute advocating for rigorous study design in biomarker validation studies. They call their design “PRoBE” for prospective –specimen collection, retrospective-blinded evaluation.” They describe methods for standardizing sample collection, defining performance criteria, and choosing cases and controls in order to avoid the most common sources of bias. For example, the commentary explains why measuring biomarker levels retrospectively can reduce bias relative to prospective measurement. The authors also describe when it is appropriate to match cases and controls instead of selecting them at random.

In a small-scale study of clinically-diagnosed ovarian cancer patients in China, Wang et al. were able to develop a seven peak SELDI-TOF-MS signature characteristic of plasma from patients with stage III/IV tumors. Using this signature they were able to correctly diagnose 9 out of 11 patients with stage I/II cancers. The size of the study is small (and the authors do not report a false-positive rate for stage I/II ovarian cancer diagnosis). However, that early-stage and late-stage plasma proteomic signatures are similar may indicate that early- and late-stage cancers are biologically similar. Many biomarker discovery studies have used clinically-diagnosed early-stage tumors as a proxy for a truly “earlier stage” of aggressive tumors. This study indicates this might be a valid proxy. However, there remains the possibility that while all clinically-diagnosed ovarian tumors are biologically similar, they are markedly different from the small, asymptomatic tumors that we need to find in early detection schemes.

Proteomics published a special issue in August dedicated to the topic "GLYCOMICS APPROACH FOR CANCER BIOMARKER DISCOVERY." Check it out here.

Kurman et al. have published a review in the American Journal of Obstetrics and Gynecology arguing that aggressive ovarian cancer is often never confined to the ovary, and thus early detection methods that focus on finding low-stage localized tumors will miss these more common, more deadly cancers. They recommend refocusing research efforts from detecting early stage ovarian cancers to detecting aggressive but small volume cancers. Read the paper and tell us if you are convinced.

Identifying genetic events that occur very early in the development of cancer could greatly facilitate the search for early detection biomarkers. For some fascinating new insights into the transition of normal cells into invasive cancer cells, check out this new paper by Majewski et al. Using a technique they have named whole organ histological and genetic mapping (WOHGM) the authors were able to correlate genomic changes with progression from normal cells to invasive bladder cancer. They mapped LOH events on a genomic scale and then narrowed in on a particular region of chromosome 13 for finer scale SNP mapping. Importantly, they were able to identify candidate “forerunner genes” that were often lost or altered in the transition from normal bladder to in situ neoplastic lesions. Two of these genes, ITM2B and P2RY5 have lowered expression in lung cancers as well.