| October 21, 2005
Medpundit nicely summarizes the recent publicity surrounding Herceptin and breast cancer.
< Previous post Previous Post
Next post > Next Post
Comments are moderated before they are published. Please read the comment policy.
Here’s a stupid question. In those patients where cancer isn’t recurring and spreading, why are they dying?
Is it because they stopped the Herceptin? The chemo and radiation made them sick or vulnerable to sickness? They all got hit by demented 93 year olds?
I felt a slight current of “Katie bar the door against an expensive treatment” by persons who have another agenda besides survival. Someone who is acutally faced with dying is going to want optimize her chances.
If the cancer is not recurring or spreading in some women who get Herceptin, why do those same women die?
I posted above.Or maybe that guy meant something else by survival rates…that the women who get recurrence and spread simply are not dying any faster than women who don’t have recurrence and spread, that is, that other treatments are keeping them alive at the same rate.
If that’s so, don’t you have to consider other things, like how miserable it might be to go through treament for recurrence and spread, even if that treatment keeps you alive as long as someone who got herceptin (who didn’t need treatment for recurrence or spread).
This is how reports can be misleading. From the abstract: “Overall survival in the two groups was not significantly different (29 deaths with trastuzumab vs. 37 with observation).” You have to understand that the study group was comprised of 1694 in the treatment arm and 1693 in the observation arm. The problem is that the chance of death is not significant so when you run the numbers and compare 97.8% survival with 98.3%, it does not constitute a big enough difference to defeat the null hypothesis of no effect so it’s reported as “not significantly different”. But look at those numbers and it is hard to conclude that there was not some effect (especially when you consider the second strudy which showed a three year reduction in deaths of 33%). An additional problem which damps the enthusiasm for herceptin is the cardiotoxicity which affected 0.5% of the treatment group.
Back of the envelope you estimate that the herceptin costs 10k per patient and you have spent about 17 million to save 8 lives, but let’s assume that eight lives are shortened and treatment for the severe cardiotoxicity is expensive (plus you don’t know about the less than svere cardiotoxicity of the rest of the treatment group) and the cost per life saved looks less attractive.
“So, it does seem to make a difference in cancer recurrence and spread, but it isn’t a dramatic life saver.” – Medpundit
ON THE CONTRARY, the study (Trastuzumab plus Adjuvant Chemotherapy…)clearly showed that there was a 33% reduction in the number of deaths in 3 years just by adding Herceptin(Trastuzumab)and giving it simultaneously to the regimen (Doxorubucin + Cyclophosphamide + Paclitazel). A reduction of 33% in mortality is highly significant. The other study(Trastuzumab after Adjuvant Chemotherapy) did not show a significant death rate 1 year later. Herceptin was given AFTER and not simultaneously. Only 26% were given a taxane (such as Paclitaxe) while everybody in the other study(one in which Herceptin was given simultaneously) received a taxane. Yes,Herceptin is a dramatic life saver if given simultaneously with adjuvant therapy and the adjuvant therapy should include a taxane. Yes, the publicity is well deserved. -Dr.EAM
“This is how reports can be misleading.”
The 2 studies differ in the timing of administration of Herceptin and the use or not use of Paclitaxel. The study seems to favor giving Herceptin simultaneously and not sequentially and giving it with Paclitaxel. The discrepancy in mortality rate may be explained by the differing methodology. _Dr.EAM
I’m not sure the results differ that much even with the differences in timing. In the one study with a median 12 month follow up, the difference was 97.8 vs. 98.3. The reduction in mortality was 24% (29 deaths vs. 37). If you extrapolate that to three years then the survival would be 95.0% vs. 93.6% which is a difference of 1.4%; in other words well within the 95% CI reported for the second study of 0.1% to 5.0%. The difference in mortality reported in the second study was 33%, but the absolute numbers (62 deaths vs. 92 deaths) were triple the first study because of the longer timeframe.
The comment that “reports can be misleading” was based on the idea that if you did not understand the way these results are reported, you might interpret the “not significantly different” phrasing to mean “no difference” or “not worth further study” when clearly there appears to be a dramatic effect.
Whether herceptin is cost effective especially given the expense and the cardiac complications is a different question. If we use the same QOLY threshold value that the USPTF used in their PSA analysis then the answer could easily be no.
The best of KevinMD.com.
Only on Facebook.
"A comprehensive and extremely useful roadmap for doctors."
-Eric Topol, MD, author of The Creative Destruction of Medicine
site by Out:think Group
Welcome to KevinMD.com, social media's leading physician voice.
Receive regular updates delivered free to your inbox.
KevinMD.com is also on Facebook, Twitter and LinkedIn.