Hyperthermia combined with radiation for the treatment of locally advanced prostate cancer

Long-term results from Dana-Farber Cancer Institute study 94-153




The authors present long-term results from a phase 2 study that assessed the efficacy of transrectal ultrasound hyperthermia plus radiation with or without androgen suppression for the treatment of locally advanced prostate cancer.


Patients with clinical T2b-T3bN0M0 disease (according to 1992 American Joint Committee on Cancer [AJCC] criteria) received radiation plus 2 transrectal ultrasound hyperthermia treatments. After the first 4 patients, 6 months of androgen suppression were allowed. The study was designed to assess absolute improvement in the 2-year disease-free survival rate compared with the short-term androgen suppression arm in Radiation Therapy Oncology Group (RTOG) study 92-02.


Thirty-seven patients received a total of 72 hyperthermia treatments. The mean cumulative equivalent minutes (CEM) T9043°C was 8.4 minutes. According to the 1992 AJCC classification, there were 19 patients with T2b tumors, 8 patients with T2c tumors, 5 patients with T3a tumors, and 5 patients with T3b tumors. The median Gleason score was 7 (range, 6-9), and the median prostate-specific antigen (PSA) level was 13.3 ng/mL (range, 2-65 ng/mL). Thirty-three patients received androgen suppression. At a median follow-up of 70 months (range, 18-110 months), the 7-year overall survival rate was 94%, and 61% of patients remained failure free (according to the American Society for Therapeutic Radiology and Oncology definition for failure free survival). The absolute rate of disease-free survival at 2 years, which was the primary study endpoint, improved significantly (84%) compared with a rate of 64% for similar patients on the 4-month androgen suppression arm of RTOG 92-02. When Phoenix criteria (PSA nadir + 2 ng/mL) were used to define biochemical failure, 89% of patients were failure free at 2 years.


Hyperthermia combined with radiation for the treatment of locally advanced prostate cancer appeared to be promising. The current results indicated that further study of hyperthermia for the treatment of prostate cancer with optimal radiation and systemic therapy is warranted. Cancer 2010. © 2010 American Cancer Society.

The benefits of hyperthermia with radiation are widely recognized. The biologic rationale for combining heat with radiation is compelling. Tumor cells that are most resistant to radiation, including those that are hypoxic, at low pH, nutritionally deprived, and in S-phase, are precisely the cells that are most sensitive to hyperthermia.1, 2 Hyperthermia also enhances the effect of radiation through the induction of apoptosis and other mechanisms of cell kill.3-5 It has been demonstrated that this theoretical benefit of hyperthermia has a meaningful clinical impact, including a survival benefit, as evidenced by results from phase 3 trials.6

Technical challenges in treatment delivery, including difficulty achieving therapeutic temperatures for deep-seated tumors, have hindered the widespread use of hyperthermia. Likewise, hyperthermia often is viewed as labor and time intensive, because many hyperthermia treatment regimens have required weekly or biweekly treatments over the course of radiation therapy.

A phase 2 study at the Dana-Farber Cancer Institute (DFCI study 94-153) was initiated to provide a preliminary assessment of the efficacy of transrectal ultrasound hyperthermia in combination with radiation and, for most patients, androgen suppression in treatment of intermediate-risk to high-risk, clinically localized prostate cancer. The use of a transrectal applicator allows for direct energy deposition into the prostate to enhance tumor heating while minimizing thermal dose to the surrounding normal tissues and organs, thus addressing the challenge of heating deep-seated pelvic organs. The use of the applicator that was assessed in the current study reportedly was feasible and safe in a previous phase 1 study.7, 8 Building on the treatment regimen applied in that earlier study, which produced promising clinical results,8 a phase 2 trial incorporated only 2 hyperthermia treatments administered within the first 4 weeks of radiation therapy spaced at least 1 week apart to minimize any lingering thermal resistance from the initial treatment. We previously reported the favorable toxicity profile of this treatment approach.9 The long-term efficacy results from that completed trial are presented here.


All patients were enrolled in a phase 2 study at DFCI, including men with clinical T2b to T3b tumors, negative lymph node status (N0), and no metastases (M0) according to 2002 American Joint Committee on Cancer (AJCC) criteria. The study was approved by the institutional review board, and all patients provided informed consent. Staging for all patients included bone scans and computed tomography scans of the abdomen and pelvis. Patients received 6660 centigray (cGy) ±5% normalized to 95% (approximately 7000 cGy; International Commission on Radiation Units and Measurements [ICRU] 38 reference dose) in 180 cGy to 200 cGy fractions. All patients received radiation to an initial field that included the prostate and seminal vesicles with a 1.5-cm margin followed by a prostate-only boost with a 1.5-cm margin. Radiation therapy was administered with a 4-field technique using ≥6-megavolt (MV) photons. Two hyperthermia treatments were administered at least 1 week apart during the first 4 weeks of radiation. After accrual of the first 4 patients, an amendment was made to the protocol to allow for the use of androgen-suppressive therapy (AST) to reflect changes in the standard of care for many patients who were eligible for the study.10-12 The recommended study regimen called for a total of 6 months of combined luteinizing hormone-releasing hormone agonist with a nonsteroidal antiandrogen, including 2 months of neoadjuvant hormone therapy before initiating radiation therapy. For patients who were receiving AST, simulation typically was performed before the initiation of treatment.

Details of the transrectal hyperthermia system have been reported previously.9 The ultrasound power was delivered from a water-cooled, 16-element, partial-cylindrical intracavitary array. Power deposition was controlled individually for each of the 16 transducers, and a closed heating/cooling system using degassed bolus water was used to control the anterior rectal wall temperature. Patients were placed in the lateral decubitus position for treatment. The placement of interstitial temperature and perfusion probes was accomplished through a transperineal route using transrectal ultrasound guidance. Three Bowman probes were placed with 1 each in the right and left lateral peripheral zones approximately midpoint in the anterior-posterior axis, and the third probe was placed in the central posterior portion of the peripheral zone. During this portion of the procedure, most patients received propofol (a short-acting, intravenous general anesthetic) supplemented with midazolam and fentanyl while spontaneous ventilation was maintained. Once the probes were paced satisfactorily within the prostate, the transrectal hyperthermia probe was introduced into the rectum. Then, patients were allowed to return to an alert state; however, at times, they received further, light intravenous sedation sufficient to communicate any pain, positional discomfort, or heat discomfort.

Next, power was applied for a minimum goal of 60 minutes at therapeutic temperature, as defined by the attainment of a temperature of 42°C by at least 1 intraprostatic temperature sensor or allowing for 10 minutes of initial heating. The objective of thermal treatment was to achieve a mean cumulative equivalent minutes (CEM) T9043°C of 10 minutes. This parameter is used to equate a range of actual temperatures achieved to a reference temperature (43°C). The temperature exceeded by 90% of the measured temperature points (T90) when given over a certain period is converted to equivalent minutes (EMs) at 43°C, as defined by Sapareto and Dewey.7 The CEM T9043°C is the summation of the EMT9043°C for each hyperthermia session over the course of treatment.

Temperature profiles were obtained for each thermocouple in 30-second intervals over the course of treatment. The maximum rectal wall temperature at any single point was limited to 40°C (19 patients), 41°C (3 patients), or 42°C (15 patients). The limitation of the rectal wall temperature to 40°C in the initial 19 patients hindered achievement of the thermal treatment goal for the majority of patients, because the applied power typically had to be reduced to keep the rectal wall temperature below 40°C. Because there was minimal rectal toxicity experienced with a rectal wall temperature limit of 40°C, the DFCI institutional review board allowed for an increase in allowable rectal wall temperature to 42°C in a step-wise manner with 3 patients treated first at 41°C. Once the session was completed, all probes were removed, and the patients received radiation within 1 hour after the completion of hyperthermia treatment.

The study was designed to have 80% power to detect a 20% absolute improvement in the 2-year disease-free survival rate (from 64% to 84%) observed on the short-term androgen suppression arm of Radiation Therapy Oncology Group (RTOG) trial 92-02. The Wilcoxon rank-sum test was used to evaluate differences in the median CEM T9043°C for different groups and differences in the median duration of treatment for different treatment sessions. Overall survival and prostate-specific antigen (PSA) failure-free survival were estimated using the Kaplan-Meier method. A 2-sided P value ≤.05 was considered statistically significant.


Thirty-seven patients received a total of 72 hyperthermia treatments between September 1997 and April 2002 on the DFCI phase 2 hyperthermia trial. The median follow-up was 70 months (range, 18-110 months), the median patient age was 64 years (range, 45-78 years. According to the 1992 AJCC classification, 19 patients had T2b tumors, 8 patients had T2c tumors, 5 patients had T3a tumors, and 5 patients had T3b tumors. The median Gleason score was 7 (range, 6-9), and the median PSA was 13.3 ng/mL (range, 2-65 ng/mL). All patients completed conformal radiation therapy with computed tomography treatment planning to a median dose of 6700 cGy (range, 6340-7200 cGy) normalized to 95%. Thirty-three patients received AST, which was initiated within 3 months before radiation therapy. All but 2 of those patients received 6 months of AST (1 patient received 9 months and another 12 months of AST). Among the first 4 study patients, who did not receive AST, only 1 patient failed at the primary study endpoint of 2 years, and an additional patient failed with longer term follow-up.

Thirty-five of 37 patients received 2 hyperthermia treatments. The median treatment duration was 62.8 minutes (range, 39-80 minutes) per treatment session and did not differ significantly between the first and second treatments. A favorable long-term toxicity profile with this treatment regimen was reported previously.9 Temperature profiles are provided in Table 1. The mean CEM T9043°C for all 37 patients was 8.4 minutes. When patients were assessed according to the allowable rectal wall temperature, those who had a rectal wall maximum of ≤40°C had a mean CEM T9043°C of 5.6 minutes versus 11.4 minutes for patients who had an allowable rectal wall temperature ≤42°C. A Wilcoxon rank-sum test indicated that the difference in median CEM T9043°C between these 2 groups (2.8 minutes vs 10.5 minutes) was significant (P = .004). A small difference in disease-free survival was noted in favor of the higher temperature group; however, it is important to note that the study was not designed to assess the impact of temperature on treatment outcome.

Table 1. Thermal Treatment Parameters
  1. EM, equivalent minutes; T9043°, temperature exceeded by 90% of the measured temperature points over a given time converted to equivalent or cumulative equivalent minutes at 43°C; CEM, cumulative equivalent minutes.

Time, min39.080.062.8
Temperature, °C   
EM T9043°C0.121.94.3
CEM T9043°C0.427.28.4

PSA failure was defined using the American Society for Therapeutic Radiology and Oncology (ASTRO) consensus definition. The Phoenix definition of PSA failure (ie, PSA nadir + 2 ng/mL) was developed after the current study was designed; and, although it was not part of the primary analysis, also was assessed to provide a basis for comparison with contemporary studies. In addition to the ASTRO definition of PSA failure, failure also was defined by clinical or pathologic evidence of local or distant disease recurrence or at the time of initiation of salvage androgen suppression regardless of PSA. At a median follow-up of 70 months (range, 18-110 months), the overall survival rate was 94% (Table 2). After 7 years of follow-up, 61% of patients were failure free according to the 1997 ASTRO consensus definition (Table 3), and 55% were failure free when we applied the PSA nadir + 2 ng/mL definition of biochemical failure (Table 4). It is noteworthy that the Kaplan-Meier curves for both overall survival and disease-free survival appeared to level off after 5 years, as illustrated in Figures 1 through 3. Three patients developed metastatic disease, and 1 of those patients died of prostate cancer 30 months after treatment. The absolute rate of disease-free survival at 2 years, which was the primary study endpoint, improved significantly with hyperthermia treatment (84%) compared with the rate of 64% for similar patients on the 4-month androgen suppression arm of RTOG 92-02 who served as the comparison group for this study. According to Phoenix criteria (PSA nadir + 2 ng/mL) for biochemical failure, 89% of patients were progression-free at 2 years.

Figure 1.

Overall survival is illustrated.

Figure 2.

This chart illustrates prostate-specific antigen freedom from failure according to the American Society for Therapeutic Radiation Oncology criteria.

Figure 3.

This chart illustrates freedom from biochemical failure according to the Phoenix criteria (prostate-specific antigen nadir + 2 ng/mL).

Table 2. Overall Survival
No. of PatientsNo. of DeathsOS Rate, %
2 Years4 Years5 Years6 Years7 Years
  1. OS indicates overall survival.

Table 3. American Society for Therapeutic Radiology and Oncology Prostate-Specific Antigen Freedom From Failure
No. of PatientsNo. of FailuresASTRO PSA FFS Rate, %
2 Years4 Years5 Years6 Years7 Years
  1. ASTRO indicates American Society for Therapeutic Radiology and Oncology; PSA, prostate-specific antigen; FFS, failure-free survival.

Table 4. Phoenix Definition (Prostate-Specific Antigen Nadir+2 ng/mL) of Freedom From Biochemical Failure
No. of PatientsNo. of FailuresPhFFS Rate, %
2 Years4 Years5 Years6 Years7 Years
  1. PhFFS indicates failure-free survival according to the Phoenix criteria.



The current study provides the longest follow-up reported to date on the use of hyperthermia as treatment for prostate cancer administered on a prospective, phase 2 trial. Previous studies demonstrated that prostate hyperthermia can be administered safely and effectively using a variety of strategies.9, 13, 14 In addition to the issues of safety and feasibility, others have demonstrated the potential benefit of using hyperthermia for the treatment of locally advanced prostate cancer.

In the earliest report we identified on outcomes with prostate hyperthermia, Anscher et al assessed locoregional therapy using an annular, phased-array radiofrequency device combined with external-beam radiation at a median dose of 6840 cGy in 21 patients who had either T2b through T4 or recurrent prostate cancer with a median Gleason score of 7 and a median PSA of 69 ng/mL. Five or 10 hyperthermia treatments were planned. In that study, the 3-year overall and disease-free survival rates were 88% and 25%, respectively. No significant increase in acute or long-term toxicity with the addition of hyperthermia to radiation was noted. The only predictor of disease recurrence was the pretreatment PSA level.15

Deger et al reported promising early results with the use of interstitial hyperthermia using implanted, self-regulating thermoseeds in 57 patients with T1 through T3 prostate cancer (95% had T2 or T3 tumors). In that trial, the median pretreatment PSA level was 11.6 ng/mL, and the median World Health Organization tumor grade was grade 2. Biochemical progression was defined according to the ASTRO consensus definition. The radiation dose was 6840 cGy in 180-cGy fractions and was administered concurrent with 6 hyperthermia treatments using implanted, 55°C Curie thermoseeds. The median follow-up was 36 months. Nine patients progressed at a median of 20 months. The median PSA was 0.55 ng/mL 2 years after therapy.16

Tilley et al reported results from 22 patients on a phase 1/2 trial who received radiation at a dose of 6840 cGy and regional hyperthermia weekly for 5 to 6 weeks. Fifteen patients in that trial had primary T3pN0M0 disease, and 7 patients had histologically confirmed local recurrences after radical prostatectomy. Five patients received short-term AST. PSA control was defined as a nadir of <1 ng/mL. PSA progression according to the ASTRO consensus was defined a PSA level >2 ng/mL, and PSA levels between 1 and 2 ng/mL were considered disease progression if the PSA level increased on 2 successive occasions. At a median follow-up of 6 years, the 6-year recurrence-free survival rate was >50% for patients with primary disease, but no long-term control was noted for patients with recurrent disease. The 6-year overall survival rate was 95% and 60% for patients with primary and recurrent disease, respectively. A clear correlation was observed between higher temperatures and thermal doses with PSA control.17

Maluta et al reported on 144 patients who had prostate cancer with T3 and T4 tumors, or T2 tumors and Gleason scores ≥7, or PSA levels ≥10 ng/mL who received conformal radiation therapy at a mean dose of 7400 cGy with hyperthermia on a phase 2 trial. Androgen deprivation was administered to 64% of patients with a high degree of variability in type and duration of treatment and lasted up to 5 years in some high-risk patients. Locoregional hyperthermia was administered weekly with the BSD 2000 Hyperthermia System (BSD Medical Corporation, Salt Lake City, Utah) up to a maximum of 5 treatments during radiation therapy. At a median follow-up of 51.7 months, the 5-year biochemical disease-free survival rate according to the ASTRO consensus definition was 49%, and the overall survival rate was 87%. Hyperthermia was tolerated well with no significant side effects noted.18

Algan et al reported the long-term results from a precursor phase 1/2 study that assessed the transrectal ultrasound hyperthermia system that was used in our current study. Twenty-six patients with high-risk prostate cancer (American Urologic Society stage C2-D1; median PSA, 29 ng/mL) received a median dose of 6800 cGy with either 1 hyperthermia treatment (9 patients) or 2 hyperthermia treatments (17 patients). Biochemical failure was defined according to the ASTRO consensus definition. The median follow-up was 71 months, and the overall and cause-specific 5-year survival rates were 73% and 79%, respectively. The median survival was 36 months, and the 5-year rate of survival with no biochemical evidence of disease (bNED) was 35% in this high-risk patient population. On multivariate analysis, a pretreatment PSA level ≤10 ng/mL was a significant predictor of bNED survival. The duration of hyperthermia therapy trended toward significance for overall survival (P = .06).8

The current results provide further support to the hypothesis that hyperthermia may be beneficial for the treatment of locally advanced prostate cancer. A significant benefit was noted with the addition of hyperthermia in this patient population compared prospectively with the study-designated control group of patients who were treated on the short-term androgen suppression arm of RTOG 92-02. Although the patients on RTOG 92-02 were similar in many ways to the patients who were eligible for the current trial and were treated within similar radiation and AST parameters, only a phase 3 study can provide conclusive evidence of efficacy.

Since the initiation of this study in the mid-1990s, much has been learned about risk stratification. More stringent eligibility criteria for identifying patients with bulky local disease who nonetheless are at relatively low risk of harboring micrometastases may aid in further defining the benefit of hyperthermia in prostate cancer, because it is unlikely that hyperthermia ultimately will benefit patients who have pre-existing, subclinical, distant disease.

Currently, it is recognized that the relatively modest radiation doses that were used in this study are associated with suboptimal tumor eradication. An argument can be made that, with current capabilities to safely dose escalate radiation, hyperthermia is not necessary for the treatment of prostate cancer. Although this assertion very well may be true for the majority of patients, the use of very tight margins required with dose escalation may not be desirable for some patients with locally advanced disease. In addition, there may be some patients who harbor relatively radioresistant tumor cells who could benefit from hyperthermia. Future developments in functional imaging and molecular profiling may lead to the targeted selection of patients for treatments that are complimentary to radiation, such as hyperthermia. Likewise, patients for whom salvage radiation therapy is being considered after primary treatment may benefit from the combined use of modest radiation doses with hyperthermia. Because there are commercially available systems to administer hyperthermia for prostate cancer, the lessons learned from the current study continue to be relevant and applicable to current clinical scenarios.

The duration of AST remains a topic of controversy. Although RTOG and European Organization for the Research and Treatment of Cancer trials have demonstrated a significant survival advantage to 2 or 3 years of AST in high-risk patients,12, 19 others have demonstrated a survival advantage for intermediate-risk and high-risk patients with 4 to 6 months of AST.20, 21 The full impact of different durations of hormone therapy, including periods intermediate between 4 to 6 months and 2 to 3 years, remain to be fully defined. Because the use of long-term AST is associated with significant morbidities, the combination of radiation, hyperthermia, and short-term AST may be an attractive alternative for some patients who have intermediate-risk or high-risk, clinically localized prostate cancer.

Apart from traditional hyperthermia geared toward radiosensitization, the use of thermal ablation for prostate cancer is gaining therapeutic momentum. There are significant challenges, however, in completely ablating the prostate without damage to the urethra, bladder, rectum, and neurovascular bundles. It is well recognized that there is a hyperthermic rim around the ablated tissue region. An attractive strategy may be to combine thermal ablation with modest doses of radiation. A safe margin for ablation near critical normal structures can be maintained by taking advantage of the hyperthermic rim to sensitize tumor in these regions to eradication with radiation. The finding of an enhanced benefit of hyperthermia with the radiation doses that were used in the current study supports such an approach. Indeed, the degree to which hyperthermia improved treatment outcome is similar to that observed with radiation dose escalation22, 23 and comes without any of the added risks associated with higher radiation doses. The benefit observed in this phase 2 study also supports the hypothesis that the thermal enhancement ratio is significant for prostate cancer and is worthy of further investigation.


Supported by National Cancer Institute grant P-01 CA31303.