Immediate or Deferred Androgen Deprivation for Patients With Prostate Cancer Not Suitable for Local Treatment With Curative Intent: European Organisation for Research and Treatment of Cancer (EORTC) Trial 30891

Urs E. Studer, Peter Whelan, Journal of Clinical Oncology, Vol 24, No 12 (April 20), 2006: pp. 1868-1876

PURPOSE: This study (EORTC 30891) attempted to demonstrate equivalent overall survival in patients with localized prostate cancer not suitable for local curative treatment treated with immediate or deferred androgen ablation.

PATIENTS AND METHODS: We randomly assigned 985 patients with newly diagnosed prostate cancer T0-4 N0-2 M0 to receive androgen deprivation either immediately (n = 493) or on symptomatic disease progression or occurrence of serious complications (n = 492).

RESULTS: Baseline characteristics were well balanced in the two groups. Median age was 73 years (range, 52 to 81). At a median follow-up of 7.8 years, 541 of 985 patients had died, mostly of prostate cancer (n = 193) or cardiovascular disease (n = 185). The overall survival hazard ratio was 1.25 favoring immediate treatment, seemingly due to fewer deaths of nonprostatic cancer causes (P = .06). The time from randomization to progression of hormone refractory disease did not differ significantly, nor did prostate-cancer specific survival. The median time to the start of deferred treatment after study entry was 7 years. In this group 126 patients (25.6%) died without ever needing treatment (44% of the deaths in this arm).

CONCLUSION: Immediate androgen deprivation resulted in a modest but statistically significant increase in overall survival but no significant difference in prostate cancer mortality or symptom-free survival. This must be weighed on an individual basis against the adverse effects of life-long androgen deprivation, which may be avoided in a substantial number of patients with a deferred treatment policy.

Several prospective randomized trials comparing external beam radiation therapy combined with concurrent androgen deprivation to radiation therapy alone and androgen deprivation at relapse showed significantly better survival for patients with concurrent androgen deprivation. Several aspects, however, suggest that the benefit of the combined treatment is most likely due to an improved local control (androgen deprived prostate cancer cells responding better to radiation) rather than to an additive systemic effect on survival. This was recently supported by D'Amico et al who showed a similar significant improvement of overall survival for patients with concurrent radio-hormonotherapy, although androgens were deprived for only 6 months.

A statistically significant overall survival advantage for immediate adjuvant hormonal treatment has been documented in the Eastern Cooperative Oncology Group study of 98 prostate cancer patients who had positive lymph nodes removed at the time of radical prostatectomy.    Similar to our study, the differences in survival are found predominantly in patients who died within 3 to 5 years after random assignment. Unfortunately, the number of patients was small and they did not report either the number of lymph nodes removed or their location. Nevertheless, these data suggest that hormonal therapy might be most effective when only minimal systemic disease is present. On the other hand, Iversen et al demonstrated that the antiandrogen bicalutamide (150 mg per day) significantly decreased the overall survival probability if given at an early stage of the disease or if used as an adjuvant after local treatment with curative intent.   Whether this is due to toxic adverse effects of this particular antiandrogen or also pertains to other forms of very early endocrine treatment remains to be seen.

A major finding of this trial is a statistically significant difference in overall survival favoring immediate androgen deprivation. It remains to be seen whether this difference will persist or even increase with longer follow-up. Indeed, it could be speculated that the actual survivors with a less aggressive prostate cancer might respond better to immediate androgen deprivation. Nevertheless, in the MRC trial the difference disappeared with longer follow-up.

The reasons why the significant survival benefit with immediate androgen deprivation was apparently due to fewer nonprostate cancer deaths and not to prostate cancer-related deaths, can only be speculated on. Until now, only 20% of all patients have died of progressive prostate cancer. In patients who are unsuitable for local treatment with curative intent, age and comorbidity play an important role; competing causes of death may mask a possible beneficial treatment effect on prostate cancer. Deferred treatment was started earlier than foreseen by the protocol in 20% of all patients, mostly due to the psychological pressure of a rising PSA or the appearance of new scintigraphic hot spots, thus potentially reducing any possible difference between the two arms. The closer the timing of immediate and deferred treatment, the less likely it becomes that a significant difference in prostate cancer related survival can be demonstrated. Despite the extensive review performed in this trial and the high agreement among reviewers of the causes of death, misclassification cannot be totally excluded as a reason for the apparent absence of a difference in prostate cancer death rates. However, the absence of a difference in the times from study entry to subjective or objective progression from hormone refractory disease after immediate or deferred androgen deprivation speaks against a selection bias in evaluating prostate cancer related mortality. Lack of statistical power is another possible explanation. Finally, it cannot be excluded that androgen ablation itself might have an impact on longevity by resulting in a generally less stressful lifestyle.

It may be a shortcoming of this trial that, despite the exclusion of patients with documented metastatic disease, patients with different prostate cancer risk profiles (including well/poorly differentiated and differences in tumor volume) were included. Although immediate treatment at the time of diagnosis appeared to be a well-defined point of time, it varies widely within the individual course of disease. Therefore, it is unlikely that the time when the disease is diagnosed is always the appropriate time to start androgen deprivation. For some patients, it may be too late to achieve a substantial beneficial effect, while many others do not require treatment at all. Indeed, at 7 years, only 49.7% of the patients on the deferred arm had started treatment (even though the majority of them had rather advanced palpable local disease) and another 25% died of other causes before ever requiring treatment. These results suggest that the deferred treatment policy could spare a substantial number of patients the known adverse effects of androgen deprivation (including negative impact on mood and cognitive function, impotence, osteoporosis, gynecomastia, headache, skin complaints).

While it must be kept in mind that the number of patients requiring a TURP because of bladder outlet obstruction was significantly higher in the deferred treatment arm, this alone would not justify immediate androgen deprivation for all patients. Also the number of patients experiencing moderate to severe pain, ureteric obstruction, or risk of pathologic fracture caused by the disease at some time during follow-up was higher in the deferred treatment arm. This is inherent to the design of the trial, which allowed the start of androgen deprivation for patients on the deferred treatment arm only on the occurrence of such symptoms or signs. Because patients were followed regularly, progressive disease was detected in time, and the risk of irreversible complications could be kept low. Of importance is that the time to first symptoms from hormone refractory disease after androgen deprivation was comparable for both treatment groups.

Additional research is underway to try to define subgroups of patients with prostate cancer that will benefit from early or, once established by ongoing trials, intermittent androgen deprivation and those patients in which this may be unnecessary over-treatment with substantial side effects that could at least be delayed for a long time. Furthermore, small but significant differences in overall survival must be carefully weighted against the possible negative impact on quality of life of lifelong androgen deprivation.

In conclusion, this study demonstrated that immediate androgen deprivation results in a significant, albeit small improvement in overall survival, but no significant difference in prostate cancer mortality or overall symptom free survival. Additional research is needed to identify possible subgroups of patients at low risk of prostate cancer progression for whom the deferred approach would be a valid treatment option