Number Crunching

My immensely supportive and very sharp friend Tania (who, by the way, surprised me the other day with a delivery of luscious Godiva chocolates … thanks, Tania!) recently asked me whether Dr. Kavanagh’s diagnosis estimates need to be updated with my MRI results.  This table summarizes what Dr. Kavanagh told me at my initial appointment with her:

50% probability – pure DCIS (noninvasive, Stage 0)
40% probability – DCIS with microinvasions (invasive cancer) present (“DCIS-MI,” Stage 1)
10% probability – DCIS with tumor(s) > 1 mm involved (Invasive Ductal Carcinoma (“IDC,” Stage ???).

While MRI turns out to be fairly competent at finding DCIS, with approx. 90% sensitivity (or yields false negatives 10% of the time), it is fantastic at finding IDC and can detect it with close to 100% sensitivity (almost no false negatives).  Compare that to mammography which has around 55% sensitivity for DCIS (false negatives for 45% of all cases!) and roughly 80% sensitivity for IDC.  The downside to its high sensitivity is that MRI also presents many false positives (hence has low specificity), resulting in unnecessary biopsies, worry, etc.  I, however, already know I have cancer and would like to discern wherever and in whichever forms it might possibly lurk.

Let’s revisit Dr. Kavanagh’s table and reassess the assigned probabilities.  To start, we can lower the probability for “DCIS with tumor(s) > 1 mm” because my MRI didn’t find any tumors > 1mm.  We can’t reduce the probability from 10% to 0% since MRI has <100% sensitivity, thus let’s call it 1%.

MRI can find lesions down to under 1mm, but does not have infinite resolution.  So while it can detect my DCIS, it cannot easily distinguish between DCIS and DCIS-MI, as any microinvasions may measure well below the 1mm mark.  Consequently, we would not want to alter the relative probabilities of “pure DCIS” and “DCIS-MI.”  Proportionally splitting the 9% we’ve recovered from our new “DCIS with tumor(s) > 1 mm” probability leaves us with all new numbers in this updated table.

55% probability – pure DCIS (noninvasive, Stage 0)
44% probability – DCIS with microinvasions (invasive cancer) present (“DCIS-MI,” Stage 1)
1% probability – DCIS with tumor(s) > 1 mm involved, (Invasive Ductal Carcinoma (“IDC,” Stage ???).

I think this is correct, but feel free to let me know if my reasoning is specious.  Tomorrow I will investigate the probability of cancer in my lymph nodes … exciting stuff!

Leave Your Comment

Your email will not be published or shared. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>