Long-term outcomes for men with high-risk prostate cancer treated definitively with external beam radiotherapy with or without androgen deprivation
Quynh-Nhu Nguyen MD,
Department of Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas
Corresponding author: Quynh Nguyen, MD, Department of Radiation Oncology, The University of Texas MD Anderson Cancer Center, 1840 Old Spanish Trail, Houston, TX 77054; Fax: (713) 563-1521; email@example.com
Men with high-risk prostate cancer are often thought to have very poor outcomes in terms of disease control and survival even after definitive treatment. However, results after external beam radiotherapy have improved significantly through dose escalation and the use of androgen deprivation therapy (ADT). This report describes long-term findings after low-dose (< 75.6 Gy) or high-dose (≥ 75.6 Gy) external beam radiation, with or without ADT.
This analysis included 741 men with high-risk prostate cancer (clinical classification ≥ T3, Gleason score ≥ 8, or prostate-specific antigen level ≥ 20 ng/mL) treated with external beam radiotherapy at a single tertiary institution from 1987 through 2004. The radiation dose ranged from 60 to 79.3 Gy (median, 70 Gy); 295 men had received ADT for ≥ 2 years, and the median follow-up time was 8.3 years.
The 5- and 10-year actuarial overall survival rates were significantly better for men treated with the higher radiation dose (no ADT plus ≥ 75.6 Gy, 87.3% and 72.0%, respectively; and ADT plus ≥ 75.6 Gy, 92.3% and 72%, respectively) (P = .0035). The corresponding 5- and 10-year biochemical failure-free survival rates were significantly better for patients treated with both ADT and higher radiation dose (82% and 77%, P < .0001). At 5 years, men who had not received ADT and had received radiation dose < 75.6 Gy had higher clinical local failure rates than those given ADT and radiation dose ≥ 75.6 Gy (24.2% versus 0%, P < .0001). The 10-year symptomatic local failure rate was only 2% for all patients.
The American Cancer Society estimates that in 2011, approximately 240,890 new cases of prostate cancer will be diagnosed and 33,720 men will die of the disease. An estimated 13% to 21% of newly diagnosed prostate cancer cases are localized and considered at high risk of local and distant relapse after primary intervention. The management strategy for high-risk prostate cancer (defined in terms of tumor size, Gleason score, and prostate-specific antigen [PSA] level) has been complicated by the perception that long-term outcome in such cases, especially biochemical disease-free survival, is poor. Historically, disease-free survival rates after external beam radiotherapy alone had been disappointing for men with locally advanced prostate cancer. Therefore, in the 1980s, attempts were made to improve the long-term outcome for men with high-risk disease by using androgen suppression therapy or 3-dimensional conformal radiotherapy, and later intensity-modulated radiotherapy, to escalate the radiation dose. The use of prolonged androgen deprivation therapy (ADT) and higher radiation doses have both led to improved outcomes,[4-6] leading to the hypothesis that this combination would likely produce additive improvements.
In addition to local and distant control, another long-term endpoint of particular interest to men with high-risk prostate cancer is the development of symptomatic recurrences that affect their quality of life. High-risk and locally advanced prostate cancer historically have been perceived as conferring a high risk of local recurrence after radiotherapy, with clinical symptoms manifesting as bladder outlet obstruction, pelvic pain, and hematuria. Investigators from our institution previously reported a rate of symptomatic local recurrence of only 8.3% at 5 years after radiation, even when patients were treated without ADT and much lower radiation doses than those used currently.
We report here significant improvements in long-term outcomes after the treatment of high-risk prostate cancer with prolonged ADT and higher radiotherapy doses. Contrary to previous assumptions, we found that very few men developed disabling local recurrences that affected their quality of life.
MATERIALS AND METHODS
This retrospective analysis, which was approved by the appropriate institutional review board, involved 741 men diagnosed with high-risk prostate cancer who received external beam radiotherapy at a single tertiary institution from 1987 through 2004. Disease was characterized as high risk if the clinical classification was ≥ T3, the Gleason score was ≥ 8, or the PSA level was ≥ 20 ng/mL. The staging work-up was standard thoughout the period of study and included computed tomography scans of the abdomen/pelvis and bone scans to rule out metastatic disease.
The median patient age at diagnosis was 68.2 years (range, 47-87 years), the median PSA level before treatment was 15.6 ng/mL (range, 1-323 ng/mL), and the median radiation dose received was 70 Gy (range, 60-79.3 Gy). The median follow-up time for all patients was 8.3 years (range, 0.13-20 years). Median follow-up times for treatment groups were as follows: for the 375 men treated without ADT and with low-dose radiation (< 75.6 Gy), 10.4 years; for the 122 men treated with ADT and < 75.6 Gy, 9.6 years; for the 71 men treated without ADT and with high-dose radiation (≥ 75.6 Gy), 10.7 years; and for the 173 patients treated with ADT and ≥ 75.6 Gy, 5.5 years. Patient and treatment characteristics are listed in Table 1.
Table 1. Patient and Treatment Characteristics by Treatment Group
Of the 741 men in this study, 295 (40%) received ADT for ≥ 2 years (range, 2.0-18 years). The ADT involved either medical castration with intramuscular leuprolide acetate injections, with or without bicalutamide, or bilateral orchiectomy. ADT was started 2 to 3 months before the initiation of radiotherapy and continued for a median of 2.9 years (range, 2-18 years) after completion of the radiotherapy. All ADT was administered for at least 2 years; no patient received less than 2 years of this therapy.
The radiotherapy techniques used were the standard of care at our institution at the time. During the early portion of the study period, treatment involved a conventional 4-field technique; subsequent techniques included 3-dimensional conformal radiation therapy and later, when computed tomography simulation and modern techniques became available, 8 coplanar-field intensity-modulated radiotherapy. The clinical target volume (CTV) included the prostate and seminal vesicles only, and did not include regional pelvic nodes. The planning target volume (PTV) included the CTV with an additional 7- to 10-mm margin superiorly, inferiorly, and laterally and a 5- to 7-mm margin posteriorly. The dose range was 60 to 79.3 Gy (median, 70 Gy) prescribed to the isocenter in earlier years and more recently to the PTV; all radiation was delivered as 1.8- to 2-Gy fractions.
Biochemical failure was defined as PSA levels exceeding 2 ng/mL over the nadir level, and local failure was defined as clinically palpable disease or biopsy-proven recurrence. Clinical failure included local, nodal, or distant failure or combinations thereof. Prostate cancer–specific disease-free survival was defined as survival without evidence of prostate cancer. Death from prostate cancer was considered an event, and patients were censored at the time of last follow-up or death from another cause. All endpoints were measured from the day of completion of the radiotherapy. Outcomes were analyzed according to use of ADT (or not) and radiation dose delivered (low or high).
All statistical analyses were performed with Stata statistical software, release 9 (StataCorp, College Station, Tex). The log-rank test was used to analyze differences between treatment groups. A Cox proportional hazards model was used for multivariate analysis.
At the time of this analysis, the institutional standard for treating high-risk prostate cancer was 2 years of ADT combined with external beam radiotherapy to a dose of at least 75.6 Gy. Therefore, we assigned patients to 1 of 4 treatment groups for the analysis: 1) No ADT with < 75.6 Gy (low dose); 2) No ADT with ≥ 75.6 Gy (high dose); 3) ADT with < 75.6 Gy; and 4) ADT with ≥ 75.6 Gy. Our goal was to evaluate whether the addition of ADT and dose escalation improved long-term outcome. Results are summarized in Table 2 and described briefly below.
Table 2. Outcomes at 5 and 10 Years According to the Use of Androgen Deprivation Therapy and High or Low Radiation Dose
No ADT, Low Dose n = 375
ADT, Low Dose n = 122
No ADT, High Dose n = 71
ADT, High Dose n = 173
Abbreviation: ADT, androgen deprivation therapy.
Low dose, radiation dose <75.6 Gy; high-dose, radiation dose >75.6 Gy. Figures in parentheses are 95% confidence intervals.
Biochemical disease-free survival rates, %
Overall survival rates, %
Prostate cancer–specific survival rates
Clinical failure–free survival rates
Local failure rates
Nodal failure rates
Distant metastasis rates, %
Biochemical (PSA-Based) Disease-Free Survival
The biochemical disease-free survival rates at 5 and 10 years after radiation were significantly different according to the use of ADT or not and higher-dose versus lower-dose radiation: both rates were highest for those given both ADT and high-dose radiation (Fig. 1A). Five-year rates ranged from a low of 45.9% (for those with no ADT and low-dose radiation) to 82.0% (for those with ADT and high-dose radiation), and 10-year rates ranged from 30.1% (no ADT and low-dose radiation) to 61.4% (no ADT and high-dose radiation) (P < .0001).
Use of ADT and high-dose radiation resulted in decreased clinical failure rates at 5 years (P < .0001) (Fig. 1B). Analyzing the specific type of failure revealed statistically significant associations for local failure (P < .0001) but not for nodal failure (P = .632) or distant metastasis (P = .081) (Table 2).
Among those patients who experienced local failure, subsequent related symptoms affecting their quality of life (urinary obstruction/retention, hydronephrosis, acute renal failure) occurred in 6.2% of patients at 10 years. The common symptomatic local failures were reported as bladder/urinary retention (2.7%), urinary frequency/obstructive symptoms (2.3%), incontinence/increased nocturia (0.4%), and hydronephrosis (0.8%) at long-term follow-up. No patient treated with ADT and high-dose radiation developed symptomatic local recurrence (Fig. 1C).
Prostate Cancer Cause-Specific Survival and Overall Survival
Prostate cancer–specific deaths were uncommon in any group within 5 years of treatment, and at 10 years after treatment only 5.5% of patients treated even without ADT and high-dose radiation had died of prostate cancer. However, after 5 years, there was a trend toward a statistical difference between the ADT and high-dose group and the low-dose groups (Fig. 2A). Although outcome for men in the no ADT and high-dose group seemed to be at least as good as that in the ADT and high-dose group, the former group was small (71 patients) and had elected to forego ADT largely because of favorable tumor characteristics.
Use of ADT and high radiation dose had a considerable positive effect on overall survival (OS) as well (Fig. 2B) at 5 years after treatment. Specifically, use of ADT with high-dose radiation produced OS rates of 92% at 5 years, compared with 82% for the low-dose and no ADT group, and the projected 10-year difference would be 72% versus 57% (Table 2).
Cause of Mortality
The causes of death for men with high-risk prostate cancer according to treatment group are summarized in Table 3. Even without the use of ADT in combination with high-dose radiation, there was a trend toward fewer prostate cancer deaths compared with that in the other treatment arms.
Abbreviations: ADT, androgen deprivation therapy; CNS, central nervous system.
Table values are given as No. of Patients (%).
The role of ADT in conjunction with external beam radiotherapy for locally advanced and high-risk prostate cancer has been investigated in several randomized trials. Trials performed by the Radiation Therapy Oncology Group (RTOG 85-31 and 86-10) showed that long-term ADT and radiotherapy improved local control and biochemical relapse–free survival compared with radiotherapy alone for patients with unfavorable-prognosis prostate cancer.[9, 10] The ensuing RTOG 92-02 trial showed that for a subset of patients with tumors of Gleason score 8 to 10, receiving ADT for 2 years after radiotherapy led to improved OS compared with radiotherapy followed by 4 months of ADT.[6, 11] Other prospective randomized studies reported by Bolla et al and D'Amico et al showed that ADT in combination with radiotherapy led to not only improved local control and decreased prostate cancer–specific mortality but also improved OS. These studies built the foundation supporting the use of long-term ADT and radiotherapy as standard therapy for locally advanced and high-risk prostate cancer.
In our experience, men with high-risk prostate cancer who received at least 2 years of ADT had significantly better 10-year rates of freedom from biochemical failure, freedom from clinical failure, and OS than did men who received no ADT, similar to the results of these previously reported randomized trials. We observed a trend toward but not a significant difference in prostate cancer–specific survival between the treatment arms. Few patients died of prostate cancer at 5 years and 10 years after radiation; the number of prostate cancer deaths were too few to likely power a statistical difference between treatment arms. Our series included a relatively small subset of patients who were not given ADT because of small disease volumes, lower PSA levels, or lower Gleason scores but were nonetheless treated with higher radiation doses. The 10-year OS and clinical failure rates for these patients were similar to those for men given ADT with radiation doses > 75.6 Gy, and this similarity probably reflected the favorable risk characteristics rather than the lack of ADT. Because these patients generally had less bulky disease, lower Gleason scores, and lower pretreatment PSA levels, they probably would have had a more favorable outcome; hence, these outcomes cannot be used as evidence to support omitting ADT for the average high-risk patient, although it does speak to the benefit of dose escalation when groups who did not receive ADT are compared.
Several randomized trials have demonstrated better PSA relapse–free survival for patients treated with higher radiation doses compared with those treated with lower doses. In one such study, rates of freedom from distant metastasis and freedom from failure were both better for men treated with 78 Gy rather than 70 Gy, with the largest benefit seen for patients with PSA levels > 10 ng/mL. In that study, at 8 years of follow-up, the use of the higher radiation dose led to superior freedom from biochemical failure rates (78% for the 78-Gy group versus 59% for the 70-Gy group, P = .004). Moreover, rates of clinical failure were significantly reduced from 15% in the 70-Gy group to 7% in the 78-Gy group (P = .014). In a more recent follow-up report, significant differences in prostate cancer–specific survival were also seen for patients with PSA > 10 ng/mL or high-risk disease treated at the higher dose level. In a nonrandomized study, Zelefsky et al[15, 16] also reported improvement in PSA relapse–free survival with the use of doses of at least 75.6 Gy; at 8 years, the PSA relapse–free survival rate among unfavorable-risk groups was 72% after delivery of 81 Gy. Hence, several trials of dose escalation have shown reductions in the risk of biochemical failure, but none of these trials has demonstrated improvement in OS or disease-specific mortality.[4, 14-17]
Evidence from other randomized trials shows clearly that radiotherapy in combination with ADT yields significantly improved clinical outcomes. This realization led to the hypothesis that use of both high-dose radiotherapy and ADT would produce maximum improvement in outcome for men with high-risk prostate cancer. Zapatero et al reported that freedom from biochemical failure rates at 5 years after treatment were 84% for high-risk patients treated with ADT and > 72 Gy versus 63% for those treated with ADT and < 72 Gy (P = .003). In this study, we found statistically significant benefits in OS, PSA disease-free survival, freedom from clinical failure, and fewer local failures when men with high-risk prostate cancer who were treated with ADT in combination with high-dose radiotherapy. The shorter median follow-up time for this group (5.5 years) reflects the relatively recent institutional practice of dose escalation for patients with localized prostate cancer. This follow-up time was shorter in the patients receiving ADT and high-dose radiation therapy. Due to the inherent limitations in our comparison to the other 3 treatment groups, we only reported our 5-year outcome for this group. Our standard practice for patients with high-risk prostate cancer is delivering ADT in combination with high-dose radiotherapy; our intention is to report our long-term outcome in this group when longer follow-up is available.
Our institutional practice does not include elective irradiation of the pelvic lymph nodes; rather, the treatment of high-risk prostate cancer has consisted of 2 years of ADT in combination with external beam radiation therapy delivered to only the prostate and seminal vesicles. Indeed, the clinical failure–free survival rate without pelvic nodal irradiation at 10 years of follow-up was 81.9% (Table 2). Moreover, no patients who received high-dose radiation and ADT experienced local failure versus 12.9% of those treated with higher doses but without ADT and 32.8% of those without ADT and lower radiotherapy doses. Our local failure rate of 6.8% for patients treated with ADT but with lower radiation doses was similar to reported local failure rates for patients treated in a randomized trial reported by Bolla et al.[12, 19] An update of their 10-year results showed a local failure rate of 4% for patients treated with long-term ADT and radiation dose of 70 Gy, indicating that long-term ADT reduced local failures even when lower radiation doses were given.
Contrary to prior belief that men with high-risk prostate cancer experience symptomatic local recurrences, we found a low 3.7% rate of symptomatic local recurrence even at 10 years among the entire cohort of patients, and no local symptomatic recurrences when patients were treated with modern radiation techniques with high radiation doses and androgen suppression therapy. At 10 years, only 2.7% of the patients experienced urinary/bladder obstructive symptoms, 0.5% of patients had hydronephrosis, 0.4% of patients had urinary frequency/urge incontinence, and 0.1% of patients reported hematuria. Thus, in the modern era, treatment consisting of 2 years of ADT in combination with high-dose radiotherapy can produce excellent outcomes at long-term follow-up, and the fear of high rates of symptomatic local recurrence is unfounded.
Our analyses indicate that long-term ADT in combination with high-dose external beam radiation therapy produces outcome results far superior to that of lingering historical perceptions. Our experience demonstrated that use of long-term (≥ 2 years) ADT in combination with radiation doses in excess of 75.6 Gy produced significant improvements in biochemical, clinical, and survival outcomes compared with previous eras; specifically, local failure rates were low and associated symptoms were uncommon. Information such as this is critical for counseling patients on treatment options.
Supported in part by Cancer Center Support (core) grant CA016672 to The University of Texas MD Anderson Cancer Center.