Paperonium

Breast cancer · BMC Med · 2023

The paper’s summary: a gene change signals faster-worsening breast cancer. Its own estimate: nearly equal risk.

1 in 100every time1 in 20the usual barabout1 in 60The summary and resultsabout4 in 5Its own ratio and range
1 in 100every time1 in 20the usual barabout1 in 60The summaryand resultsabout4 in 5Its own ratioand range

How often chance alone would produce this gap if the gene made no difference

The paper says

Of 54 patients with tumor DNA in their blood, 24 had a change in a gene called PIK3CA. Their cancer typically worsened after 12.7 months, versus 19.2 without it. The summary links the change to worse outcomes, citing p = 0.016.

Its own data show

The results print the same p = 0.016 beside its own estimate: nearly equal risk of worsening in both groups (ratio 0.93, range 0.52 to 1.69; 1 means equal), called “not statistically significant”: could easily be chance. That ratio and range give p = 0.81; the same check matches its p-values for three other comparisons.

A p-value is how often chance alone would produce a gap this big if the gene made no difference. 0.016 is about 1 time in 60, rarer than the usual bar of 1 in 20; 0.81 is about 4 times in 5. So the summary’s warning about this gene has no statistical support in the paper’s own ratio and range.

“The summary” means the paper’s abstract: the short summary printed at its top, which is the part most readers see.

Read the paper →Full technical write-up →

Potential value of ctDNA monitoring in metastatic HR + /HER2 - breast cancer: longitudinal ctDNA analysis in the phase Ib MONALEESASIA trial · BMC Med (2023) · doi:10.1186/s12916-023-03017-z