Diagnostic test performance27 Dec 2024 08:03
Just some info on sensitivity/specificity/PPV/NPV…
An updated presentation from Cizzle Bio Inc (go via the funding website and register) is highlighting 95% sensitivity and 96% NPV, which feels like it’s off new data…it doesn’t mention specificity. Those scores are really impressive, and means the test is very effective in ruling out the disease (Would recommend doing some reading on these metrics as not as straightforward as some think).
Re specificity, the last data we saw on that was 60%, which i think is fine given the nature of the cancer, but conscious the naysayers will jump on it. Better to have false positives in my view than not picking up true positives so sensitivity and NPV more important, and if you can reduce level of false positives, which they are doing, the volume of people going through further tests reduces significantly. Current CT scans can carry a 90% false positive rate so a huge improvement from the current position.
Looking back to the Western blot technique in earlier Cizzle papers, specificity was at 70% levels.
However, in a previous investor summary, where there were three different profiles (tuning the LDT to differently), they were able to bring specificity up to 87% at the expense of sensitivity. I did ask the question about whether they could run two profiles to get the best of both worlds but not heard back (unsurprisingly).
However, if I look at a recent cancer journal report, it does suggest it is plausible to have a commercial offering that covers different profiles… this was a discussion related to brain cancer…
“Using a model tuned for high sensitivity, the test was able to predict the presence of glioblastoma, the most common and aggressive brain tumour, with 100% sensitivity, and all brain tumours with 96% sensitivity, with a 45% specificity. When tuned for a higher specificity, the model yielded a sensitivity of 47% with 90% specificity.
Baker says that, by carefully selecting different signatures at each stage of the disease, they are able to detect at similar sensitivities and specificities across all stages…”
“…when you are talking about high-risk cancers, or patients with some symptoms, that may not always be appropriate. “For instance, for brain, we think an incredibly high sensitivity is needed, because the next step is a specific medical imaging test… it’s in the patient’s interest to have a higher level of false-positives than is normally acceptable. So I think it really depends upon the application.”
Given go to market activity starts in Feb…suggests it shouldn’t be long before publication of data to support the launch…
Some other diagnostic test examples…HbA1c test for diabetes…sensitivity = 32%, specificity of 94%.
Digital mammography: a sensitivity of 97%, specificity of 64.5%.
A lot depends, as you would expect, on the complexity of what you are trying to achieve…at a more straightforward level…Salignostics sal
Sign In
Follow the stocks
