SLIDES & TRANSCRIPTS
Saturday, December 14, 2002

Early Hormonal Therapy for Biochemical Failure

Martin Gleave, M.D.

And PSA detected diagnosis and treatment of early localized prostate cancer aims to do that. But this as a consequence, led to an increase in PSA detected local failures.

TOP

And this now represents the largest subgroup of advanced prostate cancer.

And again, I think semantics is an issue when discussing the timing of hormone therapy. When is early, early? And when does deferred therapy become late?

And this is especially relevant when we consider how wide the spectrum of advanced disease has become, ranging all the way from D1-D2 disease, but back to high risk localized prostate cancer. Traditionally, the debate surrounding timing of hormone therapy has focused in this area, in metastatic disease. And this really late versus later disease.

I think that the potential benefits of early therapy in oncology in general is diminished at this time, because you have lost your chance for treating at a lower tumor burden. I think that increasingly now, this group is being targeted for the potential benefits of early hormone therapy.

TOP

Again, we have to consider how it is defined. This is still being vigorously debated, but for the purposes of my discussion I'm going to define it as two consecutive rises above 0.2. And again, I don't consider treating or intervening until the PSA has increased above 0.4, and again, this is based on some of Chris Aimling's data.

Similar to radiation therapy, three consecutive rises about nadir. But similarly, I don't think that much is lost by waiting until the PSA is above 1.5.

TOP

So again, what's the nature history of PSA recurrences? The natural history is not yet well defined, but it is being defined. But like localized prostate cancer in general, the natural history is long and variable. In general, it is associated with clinical progression in most patients. And like localized prostate cancer, can be stratified into risk groups.

A number of factors to consider, both patient factors, age, co-morbidities, life expectancies, individual expectations, desire to maintain libido and sexual activity. And then there are tumor factors principally surrounding PSA kinetics, where we can help differentiate between local and systemic failures.

Again, these are post-surgery from the Pound data. I know Ed Messing is going to focus on this more, so I'll just quickly review that. A short doubling time, less than say, one year or more than one year. Timing of PSA recurrences, less than or more that two years. These will help segregate between local versus systemic. And also stratify for those at higher risk of developing clinical metastasis at less than or more than five years.

It's also important to take into consideration baseline prognostic factors, as well as pathologic stage. Is there seminal vesical invasion? Is there a positive margin, and how does that fit in with PSA kinetics?

TOP

Just one mention surrounding radiotherapy for PSA failures. Again, here the literature suggests that the lower the pre-therapy PSA level, the better. Treating at less than 1 is associated with longer and more durable control over PSA.

TOP

Again, with regards to when to initiate hormone therapy, I don't think it's well defined. Is there a best, an optimal threshold for PSA? There have been no studies in this group of patients. So, any potential benefits must be inferred from studies in other prostate cancer populations and other solid tumors.

TOP

So I'll just go through some of these that support the rationale for treating earlier, rather than deferred. And I'll start with data from pre-clinical model systems. Isaacs reported initially in the Dunning model about 20 years ago that early administration hormone therapy, plus chemotherapy was associated with best tumor control.

Again, the Dunning model is not an androgen-independent model. It is more androgen-sensitive. So again, when I found that we were going to address this in this debate, a few months ago we started designing some experiments in the Shionogi model.

TOP

This is an androgen-dependent tumor model. It grows only in male mice, and rapidly undergoes regression, followed by progression with castration. So we designed an experiment whereby we castrated groups of 10 mice Day -1 before injection; one day after injection; 3; 6; 10; and 14 days to determine at least mimicking at what time we should be administering hormone therapy.

TOP

This is unpublished data. You can see here that when mice were castrated at higher tumor burdens at day 14 after inoculation, this group was associated with the most rapid time to androgen-dependent progression. Groups castrated 6 and 10 days that were intermediate, and those that were castrated earliest were associated with the most durable control over progression.

TOP

And this is just the Kaplan-Meier curve of that data, whereby if you look at say 50 days after castration, where the vast majority of the late tumors had recurred and required sacrifice, the intermediate group, around 50 percent, and the early hormone therapy groups still had excellent tumor control. And this type of delay in tumor progression could be also be seen out much later, where traditionally these mice would have had to have been sacrificed a long time before.

So again, this is just tumor model data using an androgen-dependent model that supports early hormone therapy when tumor burden is at a minimum.

TOP

Now, we can also extrapolate from other solid tumors, for example, breast cancer. You've got adjuvant where they don't have PSA, but there are good, large phase III studies showing the benefit of early tamoxifen therapy, adjuvant tamoxifen therapy in controlling and improving survival.

There is also biological principle data, the Goldie-Coldman hypothesis, stressing that over time with genetic instability, and increasing heterogeneity will create subpopulations in tumors that are inherently resistant to the therapies that you expose the tumors to. So again, at the biologic level, this supports treating earlier at times of minimal tumor burden.

Also, in urology we tend to initially adopt a delayed approach, because of the initial VACURG trials which suggested that early versus deferred, which is really late versus later was similar. But re-analysis by Byar suggested that the cancer-specific death rates were actually lower, and any benefits of early hormone therapy were lost because of the cardiovascular side effects of estrogen, which was used at that time.

Now, I want to spend the next little while just talking about experience from combined hormone therapy and radiotherapy trials.

TOP

Again, there have been five good, large, randomized trials, all showing high recurrence rates with radiation therapy monotherapy. But when combined with hormone therapy, you get reduced local recurrence, decreased biochemical recurrences, and even prolonged or improved survival when longer-term hormone therapy is combined with the radiation therapy.

TOP

Again, I don't think this is strictly enhancing the activity of radiation therapy. It has to do with providing a systemic therapy in patients who are at greatest risk of having sub-clinical metastases at the time of treatment of their localized disease.

And this comes out from RTOG 85-31, where over 900 men were randomized to radiotherapy alone, plus deferred hormone therapy, versus radiotherapy plus immediate continuous. And again, a more recent update of this data confirmed that overall survival was improved in the subgroup at highest risk of having sub-clinical metastases.

TOP

Similar observations were seen in 92-02. Improved overall survival in high-risk patients treated with combined radiotherapy plus long-term hormone therapy. This is two years of adjuvant therapy again, now with eight years of follow-up. A survival difference was emerging in the group of high-grade patients-in other words, those at highest risk of having occult subclinical metastases at the time of their local therapy.

TOP

This is a slide given to me by Mack Roach, looking at a pooled analysis of several RTOG trials, and similar observations are seen. Those in the highest risk groups appear to have an emerging survival benefit -- this is overall survival -- when they are treated with combined long-term hormone therapy, plus radiation therapy. Again, coming to the point that earlier administration of systemic therapy in groups of patients at high risk of having metastases at the time of treatment results in improved overall survival.

TOP

This is also seen in another important and often quoted study, the Bolla series with high risk localized prostate cancer, radiation therapy plus hormone therapy, improving overall survival. And a similar and less often quoted trial by Granfors showing the same trend.

So again, these are all trials showing that early administration of hormone therapy with radiation therapy improves overall survival, as long as this hormone therapy is administered for long enough.

TOP

There is also experience from trials looking at early versus deferred hormone therapy in metastatic disease.

TOP

And for time I'll just show this one trial, which I'm sure Ed Messing will discuss. This is a trial that he led through ECOG; again, early hormone therapy versus, really, deferred hormone therapy in a group of patients with positive lymph nodes after radical prostatectomy.

Again, both the disease-free survival and the overall survival was hugely improved with the immediate administration of continuous hormone therapy. Again, this is in a group of men, many of whom have not yet developed a PSA recurrence.

TOP

So as is true for most issues in medicine, we have to be able to most appropriately select our patients. Not all patients with early and slow rises in PSA require immediate hormone therapy. Again, we want to select on the basis of PSA kinetics, and PSA doubling time, on the basis of baseline risk factors; let's stratify people into higher risk of developing metastatic disease earlier rather than later.

And we also can implement strategies to reduce the impact of the side effects of hormone therapy, such as intermittent therapy, or oral anti-androgens.

TOP

So just as a quick review, there have now been at least five phase II studies showing the feasibility of intermittent hormone therapy in groups of patients with either advanced or biochemical failures. Initially, reported by Larry Goldenberg in a mixed cohort of patients. But subsequently Grossfield, et al. from UCSF, Crook et al., while she was at Ottowa, and Nick Bruchovsky as a multi-center phase II trial across Canada.

All these were in groups with PSA failures after failed local therapy. It again showed that at least 50 percent of the time could be spent off hormone therapy, whereby especially in this trial where there were validated questionnaires showing an improved quality of life while off therapy. So this may offer a way to apply hormone therapy earlier, while balancing the impact of that early therapy on quality of life.

TOP

I'll make a plug for this trial. This is a large phase III trial run through the NCIC and SWOG. It is on the CTSU menu at the NIH. It is randomizing patients with PSA failures after failed radiation therapy, no evidence of metastasis. Their PSAs have to be above 3, and again, this was reduced.

Initially we have a PSA of 6 as the cut off, but because of poor accrual, principally in the US and through SWOG, we reduced this now down to 3, and accrual is now picking up. Nine months of hormone therapy on, and then the hormone therapy is restarted with the PSA increases between 3 and 10.

TOP

Another approach that can be used to apply early hormone therapy while balancing side effects is the use of an oral anti-androgen like casodex. The EPC trial, I'm sure you are all familiar with. It was actually a conglomerate of three separate trials in which the data was pooled. Again, it's not valid to pool this data, because of the differences in terms of the patient cohorts, and the types of treatments, and the duration of treatments.

TOP

But in essence the data is showing that for the most part it is well tolerated, except for gynecomastia and breast pain. This is not a big deal in those patients who are at very high risk of developing metastatic disease, but it is a big deal in those patients who are at a low risk. So again, it's a question of applying it.

TOP

TOP

The main endpoint again is at least an interim analysis at two years, a 50 percent reduction, and then bone scan conversion rate showing again that early hormone therapy can delay progression to metastatic disease.

TOP

Whether or not that impacts on survival, we'll have to wait probably at least another three to four years.

So just to sum up, I think that the natural history and the significance of PSA recurrence is being defined. Early hormone therapy is supported by pre-clinical data and tumor biology principles. It delays progression to bone metastases in the EPC trial and other trials, and probably provides a survival benefit. At least it's going to be inferred from other trials.

But again, like everything in medicine, we have to risk stratify, and apply this early therapy to those populations most likely to benefit from it.

Thank you.

TOP