RE: Calm B4 the storm?31 Jan 2024 17:24
Oapk20
Rob (or Allan) may be thinking about this paper:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5441127/
"For set 1, mean CIZ1b level in individuals without diagnosed tumours established a threshold that correctly classified 98% of small cell lung cancers (SCLC) and non-SCLC patients (receiver operator characteristic AUC 0.958). Within set 2, comparison of stage 1 non-SCLC with asymptomatic age-matched smokers, or individuals with benign lung nodules, correctly classified 95% of patients (AUCs 0.913, 0.905), with overall specificity of 76%, and 71% respectively. Moreover, using the mean of controls in set 1, we achieved 95% sensitivity among stage 1 non-SCLC patients in set 2 with 74% specificity, demonstrating the robustness of the classification. Thus, CIZ1b performs with clinically useful accuracy, and could potentially satisfy the unmet need for a circulating biomarker for early detection of lung cancer"
This is of course relating to lab work, and not the automated test platform, and a lot has been done since to allow an automated platform to be used, not least, the "alternative high-throughput assay formats for CIZ1B" notes as a breakthrough in the Aug 10th RNS last year.
The accuracy of the automated test is yet to be shown in the Corepath trials, but it is important to note that an LDT does not need to have a certain accuracy, it only needs to perform within its stated parameters (obviously the more accurate the better, and more clinically useful it is, but its not a case of "sorry its not 90% accurate, so no LDT accreditation is given"). I think this reduces risk and also allows it to be improved incrementally over time if necessary (with new trials)