proscapt,
I will give a direct answer to your question below, but as a background I believe that the root of your question about
the Carducci study lies in the realization that
not everyone responds in exactly the same to this (or any) treatment, based on the biology of their own unique bodies and the biology of their own unique cancers.
You may have noticed from the abstract you copied/pasted that PSADT at least
doubled (doubling PSADT is favorable, meaning PSA increases
slowed down by half or more) in 43% of the patients.
Wow!You might have also noticed that the PSA of a sub-set of that group of participants (13%) actually slowed so much that it went
down during the study.
Double WOW! Dr Nick Vogelzang (very noted PC specialist, and the doctor of at least several men here at HW) called the results “
dramatic.” (He reported that he recommends pomegranate extract or juice “a lot” for men in his practice with rising PSA.)
Clearly, for the rest of the participants, their PSADT either improved at a lower rate, or their PSA continued to progress at the predicted rate for the duration of the study (up to 18 months) or their results were somewhere in between these. The sentence you have highlighted indicates that out of this group of participants, many decided to no-longer forego any/all additional therapies (as required by the study), and so they dropped out of the study to pursue further treatment.
proscapt, if I further “read between the lines” of your question, I might proactively point out that in an “intent-to-treat” (ITT) analysis, the results of
ALL participants are included in the results of the original groups to which they were randomly assigned, regardless of subsequent withdrawal from the program or deviation from the study protocol. So the PSA changes of all those 42% who dropped out (many because of rising PSA) were also
included in the final median increase of PSADT from 11.9 to 18.5 months. Digest that, and you can see that for many (not all) men the results were indeed "
dramatic!"
For completeness, one should note that different types of studies, for example, a “per-protocol” analysis, would have
excluded the results of those that dropped out, but the Carducci study was an ITT study which
included all results.
For further completeness, also see my descript
ion of "median"
HERE; skip down to about
the 5th post in this thread.
Does this help…?