A paper in the October issue of the Journal of Urology offers an interesting analysis of the increase in median overall survival of men newly diagnosed with metastatic prostate cancer, not previously treated with androgen deprivation therapy (ADT), and enrolling in major clinical trials between 1985 and 2009.
This paper by Tangen et al. very nicely illustrates that there has been a significant increase in the median survival of men meeting a similar set of trial eligibility criteria (evident metastases, ADT-naive, etc.) for three large, intergroup clinical trials coordinated by the Southwest Oncology Group (SWOG) since before the widespread introduction of PSA testing (i.e., 1985) to the present day. The three trials are:
- SWOG 8494, which enrolled 617 patients (between March 18, 1985 and April 2, 1986), all of whom were treated with an LHRH agonist and were then randomized either to the antiandrogen flutamide or to a placebo. At the time of progression, patients receiving a placebo could elect to “cross over” to receive flutamide.
- SWOG 8894 enrolled 1.387 patients (between December 15, 1989 and September 15, 1994), all of whom underwent a bilateral orchiectomy and were then randomized to flutamide or a placebo. Again, at progression, patients receiving a placebo could elect to “cross over” to receive flutamide.
- SWOG 9346 was the trial of intermittent vs. continuous ADT reported earlier this year at the ASCO annual meeting that enrolled 3,040 patients (between 1995 and 2009), of whom 1,535 were randomized to intermittent or to continuous ADT.
Now the statistical analysis of the data from these three trials is complex because of the fact that there are variations in eligibility criteria and many patients were missing specific data needed to assure reasonable comparisons. So, we will simply present the results of the analysis in the same way as the authors do. Those who really want to “get into the weeds” will need to get a copy of the entire text of this article (which we have read):
- The median overall survival of patients in SWOG 8494 was 31 months.
- Non-African American patients in this trial had a median overall survival of 33 months.
- African American patients in this trial had a median overall survival of 27 months.
- The median overall survival of patients in SWOG 8894 was 33 months.
- Non-African American patients in this trial had a median overall survival of 35 months.
- African American patients in this trial had a median overall survival of 27 months.
- The median overall survival of patients in SWOG 8494 was 49 months.
- Non-African American patients in this trial had a median overall survival of 49 months.
- African American patients in this trial had a median overall survival of 48 months.
There are two very clear pieces of information evident from these results:
- The median overall survival of men being diagnosed with M+ disease and enrolled in SWOG 9346 was 16 to 18 months longer than that of comparable patients enrolled in the two earlier trials.
- The median overall survival of African Americans enrolled in SWOG 9346 was almost exactly comparable to that of non-African Americans, which had not been true of the two earlier trials.
It is certainly arguable that the increase in the use of PSA testing and the resulting “stage shifting” of the disease may have something to do with this increase in overall survival. However, the authors are very careful to point out that they cannot draw any absolute conclusions based on the available data. On the other hand, they do state the hypothesis that,
… the survival improvement for all patients, particularly [African America] men, is based on a significant shift to less extensive disease.
The precise reasons why we are seeing this (very gratifying) improvement in overall survival of men diagnosed with M+ disease are going to be much harder to draw out, however. Are we identifying men with metastatic disease earlier and with lower disease burden than we did in the 1980s? Almost certainly we are. Is PSA testing and monitoring contributing to this? Probably. Are improved forms of imaging test assisting in this? Again, probably. At the other end of the scale, we also need to remember that in the time frame under discussion we saw the approval and testing of multiple new products that have clearly improved survival of men who become castration-resistant (from docetaxel to radium-223).
Filed under: Management, Treatment Tagged: | improvment, metastasis, overall, survival
Search for new and
ongoing trials on the
CTAG PCa web site
Both cohorts qualified for the label “evident metastatic disease”, but the meaning of this label changed significantly from 1985 to 1995 to 2005.
To compare apples to apples, the latter two cohorts should each have been fractionated into subgroups “disease that would have been labeled M1 in 1985|1995.”
Then a valid comparison could be done: (1) Among three “M1-by-1985 standards” groups; and (2) Between the two “M1-by-1995 standards” groups.
(This is the latest installment in a series of rants about the need to providing correct labels and classifications whenever possible, instead of perpetuating labels that are obviously prone to misunderstanding. This error occurs when two incomparable things are compared as if they were the same simply because they share a common [bad] label — in this case, “M1″.)
Paul:
That’s a nice idea … but I’m not actually sure that it is even possible — because it would depend on whether the same categories of data had been collected with the same level of accuracy in all three trials.
Paul:
Actually … now that I think about it … the term M1 wasn’t even used in prostate cancer before about 1992. Men with metastatic disease were being classified according to the Jewett-Whitmore system (as D1 or D2), and so I have to assume that the authors of the current paper did actually try to ensure that they were comparing apples to apples to whatever extent they felt able to do this.