How long should you stay on hormones when you get radiotherapy?

We have just come across a retrospective analysis of a subset of patients originally enrolled in the RTOG 85-31 clinical trial and suggesting that patients who received ≤ 5 years of hormone therapy in association with radiation did worse clinically than those who received > 5 years of adjuvant hormone therapy.

Now given the stature of the authors of this paper by Souhami et al., we have little doubt that the analysis is accurate. (We should be clear that one of the authors — Dr. Sandler — is a respected member of the Scientific Advisory Board to The “New” Prostate Cancer InfoLink.) However, there are things that worry us about the possible implications of this analysis.

Radiation Therapy Oncology Group 85-31 was a randomized trial of androgen suppression for life for patients with locally advanced prostate cancer. The original trial was started in 1987 (now 22 years ago) and the initial report of the data from this trial was published 12 years ago (i.e., in 1997) by Pilepich et al. The study originally enrolled nearly 1,000 patients, and data from 945 of these patients was suitable for analysis at the time of the initial report. Of these 945 patients, 477 received radiation + goserelin acetate (theoretically for life) and 468 were placed on observation after hormone therapy. (In other words, this was not even a randomized trial!) At the time of the initial report, the 5-year survival rate (for the entire population) was 75 percent on the adjuvant hromone therapy arm versus 71 percent on the observation arm (P = 0.52). However, in patients with centrally reviewed tumors with a Gleason score of 8 to 10, the difference in actuarial 5-year survival (66 percent on the adjuvant goserelin arm vs 55 percent on the observation arm) reached statistical significance (P = 0.03).

Over time, many of the patients in RTOG 85-31 who were originally randomized to receive hormone therapy “for life” stopped taking the hormones — probably because they had no evidence of recurrence and/or the side effects of hormone therapy were too difficult for them to deal with. In their retrospective analysis of the data from this trial, what Souhami et al. have tried to do is to correlate time on adjuvant hormonal therapy with clinical outcome among the patients who prematurely discontinued hormonal therapy.

Apparently there was a total of 189 analyzable patients (39.6 percent of the originnal 477 randomized to adjuvant hormone therapy). These patients were divided in three subsets based on the duration of hormone therapy:

  • Those receiving hormone therapy for ≤ 1 year
  • Those receiving hormone therapy for > 1  and ≤ 5 years, and
  • Those treated with hormone therapy for > 5 years.

The authors then assessed rates of overall survival, disease-free survival (DFS), cause-specific mortality, local failure (LF), and distant metastasis (DM) for the patients by subset, and noted the following results:

  • The median follow-up for the surviving patients was 9.6 years.
  • The median overall duration of adjuvant hormonal therapy was only 2.2 years.
  • The group of patients who remained on hormonal therapy > 5 years showed a significantly improved overall survival and DFS and fewer DMs than either of the groups of patients who stayed on therapy for ≤ 5 years.
  • After adjustment for age, radical prostatectomy, nodal status, Gleason score, and stage variables, the group of patients who remained on hormonal therapy for > 5 years still demonstrated superior overall survival and DFS than either of the groups of patients who stayed on ntherapy for ≤ 5 years.

Based on this analysis, the authors conclude that adjuvant hormonal therapy following radiotherapy for a period of > 5 years appears to be significantly associated with improvements in most outcomes. They go on say that a duration of horm one therapy of ≤ 5 years may have a detrimental effect on patients with locally advanced prostate cancer.

Now the first and most important question raised by this analysis is whether the patients enrolled in RTOG 85-31 are in any way representative of patients receiving treatment in the USA for locally advanced prostate cancer today. RTOG 85-31 enrolled patients who almost certainly never received a PSA test, had been diagnosed with a DRE, and in many cases would have had positive lymph nodes and significant micrometastatic disease at the time of enrollment. Finding 1,000 patients to enroll in such a trial today might prove quite challenging in the USA!

The second issue is whether the small percentage of patients who were elegible for analysis is really representative of the original enrollees in RTOG 85-31.

There is now a considerable amount of data available in more “modern” patients suggesting that short periods of neoadjuvant and adjuvant therapy are sufficient to control locally advanced prostate cancer in combination with the higher doses of radiation that are commonly give today. Although the authors clearly state that it would be necessary to complete a new randomized clinical trial to study the appopriate duration of hormone therapy, The “New” Prostate Cancer InfoLink wonders whether such a study is even feasible today, given the wide variety of clincial options that are available for a patient with a rising PSA after definitive surgical treatment. Patients would now have to be stratified by the D’Amico PSA doubling time criteria (into low, intermediate, and high risk groups) in addition to the need to consider (at least) three or four different types and periods of hormone therapy.

At the end of the day, there is an art to medical care. It may not be feasible, sometimes, togenerate the type of level 1 evidence that one would like in order to determine whether therapeutic option A really is significantly better that therapeutic options B, C, and D. This may be one of these cases. We know far more today about the adverse effects of hormone therapy than anyone knew in 1985 when this study was originally designed, and one of the things we know is that living on hormone therapy for life is extraordinarily difficult for many men and can even be a cause of death in itself as a consequece of the complications of metabolic syndrome.

Would it be nice to know exactly what the optimal form and the optimal duration of neoadjuvant and adjuvant hormone therapy should be in men who need radiation for locally advanced disease? Yes, of course it would. Do we have another 15-20 years to complete the study that might offer us these data? Probably not. The “New” Prostate Cancer InfoLink would like to believe that, in that 15-20 year timeframe, we can get better at treating earlier stage disease and later stages of disease, such that the need for this type of combination of radiotherapy and hormone therapy becomes a dinosaur — along with all its complications.

%d bloggers like this: