"Holygrail of cancer research":doctors positive about early detection blood test

[ignhelper]

https://www.theguardian.com/society/2018/jun/01/doctors-welcome-possible-holy-grail-of-cancer-research?CMP=Share_iOSApp_Other

Good, cancer is terrible and nobody deserves to suffer with it

[Headhumpy]

The test was most accurate for diagnosing pancreatic, ovarian, liver and gallbladder cancers, correctly finding the diseases in at least four out of five patients.

The blood test found lymphoma and myeloma with slightly less accuracy, at 77% and 73%, and bowel cancer in two out of three patients. Lung cancer was detected in 59% of patients. Head and neck cancer was detected in 56% of patients.

Those are some awful numbers, and no word on the false positive rate either. Derek Lowe wrote a post on this topic earlier this year when another paper about liquid biopsies came out. We've got a long way to go before liquid biopsies can actually be used in the clinic. Even if they become a routine diagnostic tool, their usefulness is doubtful since overdiagnosis and overtreatment is very much a thing with cancer (see, for example, breast cancer screening).

[SKS]

Holy Grail so holy it hit 75% with a sample sizes as strong as 10 over a testing period that might very well have missed the other 25% undetected, undeveloped cancers. Could also just be a statistical anomaly at those weak sizes.

Did I miss it or did the article also not mention false positive rates?