Five-year quality of life in patients with high-risk localized prostate cancer treated with external beam radiotherapy alone versus external beam radiotherapy with high-dose-rate brachytherapy boost: a prospective multicenter study

Purpose

Radical radiotherapy is a well-established treatment option for clinically localized and locally advanced prostate cancer (PCa) [1]. Technological advances in radiotherapy have been rapidly incorporated into clinical practice to reduce treatment-related morbidity and improve oncological outcomes [2,3]. Several randomized trials have demonstrated that dose-escalated radiotherapy-external beam radiotherapy (EBRT) or EBRT plus high-dose-rate (HDR) brachytherapy (BT) boost, improve both local and biochemical control in intermediate- and high-risk PCa [3,4,5,6], despite an increased risk of late genitourinary and gastrointestinal toxicities. The addition of HDR-BT boost allows for highly conformal dose escalation and greater sparing of the surrounding healthy organs [5].
All of the main treatments for PCa have side effects, and radical prostatectomy (RP) seems to produce greater urinary incontinence. BT is associated with an increase in urinary irritative/obstructive symptoms, while EBRT has a greater impact on bowel-related indications [7,8,9]. In selected patients, BT is an alternative to RP, limiting the risk of urinary incontinence as well as the potential impact of sexual dysfunction on health-related quality of life (HRQL) [7]. In this context, the main objective of this study was to determine whether EBRT alone or EBRT + HDR-BT was associated with better HRQL outcomes in patients with high-risk PCa, five years after the treatment. Secondary outcome measures included biochemical relapse-free survival (BRFS), overall survival (OS), and cancer-specific survival (CSS) at 5 years.

Material and methods

Design and study population

This was a prospective, multicenter study of patients with high-risk PCa, treated at participating centers from 2004 to 2006, and followed for ≥ 5 years after the treatment. The study protocol was approved by clinical research ethics committees of the six participating hospitals. Written informed consent was obtained from all patients. Staging and risk group classification were performed using a TNM staging system of the American Joint Committee on Cancer [10] and a risk group classification system developed by D’Amico et al. [11]. Inclusion criteria were biopsy-proven high-risk PCa (≥ stage T2c, prostate specific antigen [PSA] > 20 ng/ml, or Gleason > 7) without previous transurethral resection.

Clinical evaluation

Serum PSA levels were measured at all follow-up visits, performed every six months for the first two years and annually thereafter. Biochemical failure was defined as an increase in PSA levels ≥ 2 ng/ml above the nadir after radiotherapy, in accordance with updated recommendations from the Radiation Therapy Oncology Group-ASTRO Phoenix consensus panel [12].

Treatment

Treatment decisions were made jointly by patients and physicians. All patients received neoadjuvant androgen deprivation therapy (ADT). Adjuvant ADT was prescribed for 2-3 years, in accordance with clinical guidelines [13]. In most cases, patients received an anti­androgen combined with LHRH analogues. In all cases, EBRT was performed with the patient in supine position, with legs and feet immobilized. All patients underwent a computed tomography (CT) scan in the treatment position. The results of this scan were entered into three-dimensional (3D) treatment planning system and used to contour the prostate, vesicles, bladder, and rectum. External beam clinical target volume (CTV) was defined on CT imaging to cover the prostate gland and seminal vesicles with a 1 cm margin, except posteriorly, where the margin was reduced to 5 mm to create planning target volume (PTV). Custom blocking with multileaf collimators was designed using a beam’s-eye-view, and additional margins were adjusted to provide minimum dose of 95% to the prostate PTV. Risk organ constraints included the femoral heads (mean dose ≤ 45 Gy) and bladder/rectum (V70 < 25%; V60 preferably < 40%, maximum 60%; V40< 60%, maximum 80%) [14]. Off-line setup control was assessed weekly by comparing orthogonal portal images with the corresponding digitally-reconstructed radiographs.
External beam radiation therapy was performed using 3D conformal technique by photons from 15 to 18 MV isocentric conformal fields. In most patients, a six-field technique without pelvic irradiation was used. The EBRT alone group received daily fractions of 1.8/2 Gy, 5 days per week, for a mean total dose of 73 Gy to the PTV. For the HDR-BT boost, the patients were placed in lithotomy position under spinal anesthesia. A needle guidance template was attached to an ultrasound probe close to the perineum, and the needles were inserted under transrectal ultrasound guidance. Needle depth was determined by direct visualization on ultrasound and fluoroscopy. A CT scan was performed for volume delineation of the prostate and risk organs. The dose constraints for the HDR-BT boost after EBRT were as follows: PTV with V100 ≥ 98%, V150 ≤ 50%, and 105% < D90 < 115%. The constraints for the rectum were: D2cc ≤ 75% and Dmax < 100%; and for the urethra: D2% < 120%. The details of radiotherapy technique performed at our center have been published elsewhere [15]. The HDR-BT boost was administered using a temporary iridium-192 implant in one or two fractions separated by 6 hours: 21 patients (23.5%) received one fraction (20/2: 9 Gy and 1/21: 9.5 Gy) and 68 patients (76.4%) received two fractions (dose range, 6-11.5 Gy). For the patients treated with EBRT + HDR-BT, the mean EBRT dose was 51.08 Gy and the mean BT dose was 17 Gy.

Assessment of HRQL

Health-related quality of life questionnaires were administered telephonically by trained interviewers with wide expertise on this population, before treatment and during follow-up at 1, 3, 6, 12, 24, 36, 48, and 60 months after the treatment. HRQL was evaluated using validated Spanish-language versions of the following instruments: Short form-36 (SF-36), version 2 [16], Functional Assessment of Cancer Therapy – general and prostate (FACT-G and FACT-P) [17], Expanded Prostate Cancer Index composite (EPIC) [18], and International Prostate Symptom Score (IPSS) [19]. The SF-36 contains 36 items with two summary scores: the physical and mental component summary (PCS and MCS), with scores ranging from 0 to 100 on each dimension. The FACT-G (version 4.0) contains 27 items in four dimensions, measuring physical, social, emotional, and functional well-being. The prostate module (FACT-P) is specific for PCa patients and includes 12 questions about urinary symptoms, bowel and sexual function, and pain. Scores range from 0 to 108 on the FACT-G and from 0-48 on the FACT-P. The 50-item EPIC instrument evaluates four domains (urinary, bowel, sexual, and hormonal), with two urinary scales that distinguish between irritative/obstructive symptoms and incontinence. The final score ranges from 0 to 100. Higher scores indicate better quality of life (QoL) in all these questionnaires. The IPSS assesses urinary symptoms, with one question about HRQL. The total score range from 0 to 35, with higher score indicating worse symptoms.

Sample size

Sample size calculations were based on expected between-group differences on change of HRQL scores as being the principal objective of the present study. It was calculated that a total of 129 patients would be required to detect a difference in change between treatments’ groups of 0.5 standard deviation in any HRQL score, given a statistical power of at least 80% at a significance level of 5%.

Statistical analysis

Categorical variables were expressed as frequencies and percentages, and continuous variables as mean and standard deviation (SD). Differences in the distribution of variables between the study groups were compared using chi-square (χ2) test or one-way analysis of variance (ANOVA) with post-hoc Tukey’s procedure, whenever appropriate. To assess HRQL changes over time, while accounting for correlation among repeated measures, separate generalized estimating equation (GEE) models were constructed for each specific HRQL score (FACT-P and EPIC) and for the generic ones (SF-36 and FACT-G), all included as dependent variables. Time was included in the model as a categorical variable, and interactions between treatment and time were considered to test differences in trends among treatment groups, after adjusting for age, risk group, and pre-treatment prostate volume. Differences in BRFS, OS, and CSS at 5 years were analyzed with Kaplan-Meier method and log-rank test. Statistical significance was set at p < 0.05. SPSS, v.22.0 and SAS/STAT®, v.9.4. were used to perform data analyses.

Results

The study population was comprised of 129 patients, 41 treated with EBTR alone and 88 with EBRT + HDR-BT. All patients received neoadjuvant ADT. Adjuvant ADT was prescribed for 2-3 years in accordance with clinical guidelines [13], although the final date of ADT administration was not registered. Table 1 shows patients’ clinical characteristics at baseline, mean pre-treatment HRQL scores, and response rate during follow-up. The only statistically significant differences between treatment groups at baseline were a slightly higher mean number of comorbidities in the EBRT group (3.1 with EBRT vs. 2.5 with EBRT + HDR-BT, p = 0.043) and a higher SF-36 PCS score in the EBRT + HDR-BT group (50.8 vs. 53, p = 0.04). Overall, the response rate to the HRQL questionnaires during the study period was high in both study groups: at 5-years of follow-up, the response rate was 100% (EBRT) and 97.1% (EBRT-BT), without significant between-group differences (p = 0.354).
Mean changes in QoL scores from baseline to 5-year follow-up are shown in Table 2. Compared with patients treated with EBRT + HDR-BT, the EBRT alone group had significantly lower (worse) hormonal scores (–10.6 vs. –2.4, respectively, p = 0.028). Tables 3 and 4 present the results from GEE models constructed for the specific and generic HRQL scores, respectively. For the EBRT alone group, statistically significant differences from baseline were only found 5 years after the treatment for FACT-P (β = –2.4), EPIC bowel and hormonal (β = –3.1 and β = –10.4, respectively), among the specific HRQL scores evaluated. No statistically significant differences were found between the two treatment groups, with the exception of a lower impact on the hormonal domain at the EBRT + HDR-BT group.
At 5-years post-treatment, the EBRT alone group showed greater deterioration in SF-36 PCS (β = –9.8 at years) and in SF-36 MCS (β = –7.1) compared to the combined group. The same pattern was demonstrated by almost all FACT-G domains (Table 3). Figures 1 and 2 show the differences between groups in the mean HRQL scores during 5-year follow-up. We did not observe any substantial short- or long-term differences between the groups. Moreover, no statistically significant differences in BRFS, OS, and CSS between the study groups were observed (Figures 3-5).

Discussion

Health-related quality of life is an important outcome measure in patients with prostate cancer. However, long-term differences in HRQL outcomes in these patients have rarely been reported. Most of the available data on HRQL comes from studies, which have assessed the impact of treatment (BT, RP, and EBRT) on low- and intermediate-risk patients [20,21,22]. Few studies have compared EBRT alone to EBRT + HDR-BT in high-risk PCa patients in terms of HRQL [23,24].
In our study, we did not observe any significant between-group differences in mean changes in the HRQL questionnaire scores from baseline to 5-year follow-up. However, we did notice a significant difference in the EPIC hormonal domains. Although both groups experienced substantial deterioration in this domain, the EBRT alone group had significantly lower (worse) hormonal scores at 5-year follow-up. Moreover, the GEE model revealed greater worsening mainly in the FACT-P, FACT-G, SF-36, and bowel domain in the EBRT group in the fifth-year post-treatment. The greater deterioration of hormonal and sexual scores was most marked until the second year after the treatment. After this time point, there was a trend towards an improvement in these scores, probably due to the finalization of ADT and the consequent improvement in ADT-related side effects. This finding could be due to the slightly better (p = 0.135) baseline EPIC hormonal summary scores in the EBRT group.
Overall, findings of this study suggest that EBRT alone or combined with HDR-BT appear to have similar effects on HRQL. However, in our sample, patients in the BT boost group had fewer comorbidities at baseline, which could partially explain why we observed no major between-group differences. Indeed, this probably also explains why they were eligible for HDR-BT boost. Further studies with longer series and longer follow-up would provide more information about the impact of these radiotherapy modalities on HRQL. Dose escalation with EBRT followed by brachytherapy boost have been shown to improve BRFS [5,15,25]. In the ASCENDE-RT trial [26], patients were randomized to dose-escalated EBRT (78 Gy) or EBRT plus low-dose-rate BT boost. The patients in the boost group were twice as likely to be free of biochemical failure at a median of 6.5 years of follow-up, without significant differences in OS. Theoretically, dose escalation with HDR allows for an increase in biologically-effective dose, improving tumor control and sparing organs at risk [22]. However, in our series, we found no statistical significance between-group differences in BRFS, OS, or CSS. Moreover, our BRFS results differ from those reported in randomized trials [22,23,25,26], probably because of the small number of patients. Certainly, the lack of significant differences in survival outcomes in our study may be due to the limited sample size, which was powered to detect differences in the main study variable (HRQL), but not for survival outcomes.
Other authors, such as Ferrer et al. [18], have reported that the addition of ADT causes temporary deterioration in some HRQL domains, a finding that is consistent with the worsening observed in our patients in the EPIC hormonal domain. Those authors observed that ADT was associated with worse results related to vitality, hormonal function, and sexuality. Only a few studies with long follow-up have been performed to assess HRQL in patients with high-risk PCa [22,27]. In this group of patients, it is necessary to achieve the most appropriate treatment to improve local control with the lowest toxicity and the best possible HRQL. The main limitations of our study are the relatively small study population, the differences in group size, and the lack of randomization. Another limitation is the use of conformal 3D-RT rather than more advanced techniques, such as intensity modulated radiation therapy (IMRT) or volumetric modulated arc therapy (VMAT). However, during the study period (2004-2006), neither of those techniques was available in the participating hospitals. Nevertheless, given the similar clinical characteristics at baseline in the two groups, the comparison can be considered valid. Study strengths include the prospective design and the fact that the same two trained interviewers administered all HRQL questionnaires during the entire period of study.

Conclusions

In the present study, EBRT administered alone or in combination with HDR-BT boost had a similar impact on HRQL in patients with high-risk localized PCa over 5-year follow-up. However, patients in the EBRT alone group experienced significantly greater worsening in hormonal symptoms (EPIC questionnaire) at 5-year follow-up, perhaps due to the higher EPIC hormonal summary scores at baseline in this group. Longer follow-up would be needed to minimize the effects of ADT on HRQL and to determine the best treatment in terms of HRQL.

Acknowledgments

The authors would like to thank Marta Pulido, MD, and Bradley Londres for professional English language editing.

Disclosure

The authors report no conflict of interest.

References