Significance of androgen-deprivation therapy for intermediate- and high-risk prostate cancer treated with high-dose radiotherapy: A literature review

Literature search

A qualitative systematic review was conducted in 2022 based on the Minds guidelines.¹³ We performed a comprehensive literature search of two databases (PubMed and ICHUSHI) for all relevant published studies between September 2010 and August 2020, using a search formula created based on three key studies selected by four reviewers (RA, HI, SS, and MK) and the following ten keywords: “prostate cancer,” “radiotherapy,” “androgen deprivation therapy,” “high risk,” “intermediate risk,” “IMRT,” “brachytherapy,” “proton therapy,” “carbon-ion radiotherapy,” and “high-dose.” We also performed a search to include relevant papers. Given the marked advances in this area, papers published prior to August 2010 were not included.

Methods of study selection, quality assessment, and data extraction

As primary screening, two pairs of reviewers inspected the titles and abstracts of studies identified by the search. Half of the studies were independently reviewed by one of the pairs, and the other half were similarly reviewed by the other pair. As secondary screening, full papers of the studies selected in primary screening were inspected. Half of the selected studies were independently reviewed by one pair, and the other half were done so by the other pair. In addition, we also performed hand-based searching of relevant manuscripts. Studies meeting the following criteria were selected: (1) main subjects comprised intermediate- and high-risk cases; (2) overall survival (OS), PCa-specific survival, progression-free survival, BF-free survival, toxicities due to ADT, result of quality of life (QOL) analysis, and late toxicities due to RT were included in the results; (3) comparison of these clinical outcomes according to the duration of hormonal therapy were included in the analyses; (4) total doses exceeding 70 Gy in conventional fractions were prescribed for the prostate in EBRT studies. Any disagreements during primary and secondary screening were resolved through discussion and reaching a consensus.

After the second screening, the results of each paper were organized based on information on, “Patients,” “Interventions,” “Comparisons,” “Control,” and “Outcomes” from each paper and evaluated in the qualitative synthesis.

Evidence analysis

Figure 1 shows a process of literature search. The systematic literature searches yielded 498 references, which included three studies selected by hand search: 467 from PubMed and 31 from ICHUSHI. Among them, 403 papers were excluded based on the title and abstract (the primary screening process). After review of the full text of the remaining 95 papers, 54 were excluded (secondary screening process). Finally, 41 studies were included in the systematic review.^14-53 The majority were retrospective cohort studies (N = 34). Regarding RCTs, five papers on four RCTs were included.

Impact on overall survival

The phase III PMH 9907 trial evaluated the effect of adding androgen receptor blocker monotherapy (bicalutamide, 150 mg once daily) for 5 months predominantly among IR PCa patients (94.6% of study population) treated with definitive EBRT (Table 1).¹⁷ In this trial, 75.6 to 79.8 Gy in 42 fractions or 78 Gy in 39 fractions were prescribed for the prostate. OS was not significantly different between the two groups: 82% (95% confidence interval [CI]: 75%–90%) in the EBRT plus bicalutamide group versus 86% (95%CI: 80%–94%) in the EBRT-alone group at 9 years, p = 0.37. The other phase III DART01/05 GICOR trial¹⁴ compared the outcomes of 4-month ADT (2-month neoadjuvant ADT [NA-ADT] using combined androgen blockade [CAB] followed by 2-month concomitant ADT) (short-term ADT) with those of 28-month ADT (same treatment followed by 24-month adjuvant ADT [A-ADT]) (long-term ADT) in combination with high-dose EBRT (76 Gy in 38 fractions) among patients with clinical stage T1c–T3bN0M0 PCa with IR and HR factors. After a median follow-up of 63 months, OS was significantly better in the long-term ADT group. Specifically, the 5-year OS rates were 95% (95%CI: 93%–97%) in the long-term ADT group vs. 86% (95%CI: 83%–89%) in the short-term ADT group (hazard ratio [HR]: 2.48 [95%CI: 1.31–4.68], p = 0.009).

The impact of adding ADT on survival outcomes was different between the IR PCa and HR PCa populations. In subgroup analysis of the aforementioned phase 3 RCT,¹⁴ no significant difference in OS was observed among IR PCa populations (p = 0.318) while OS was significantly better in the long-term ADT group among HR PCa populations (p = 0.015). According to the results from several retrospective studies, adding ADT consistently failed to improve survival outcomes among IR PCa populations, irrespective of the ADT duration.^18-23 On the other hand, its impact on HR PCa populations was not consistent among results from retrospective studies, including this systematic review. According to a Japanese multi-institutional retrospective analysis of HDR brachytherapy combined with EBRT,²⁰ OS was significantly more favorable in patients who received NA-ADT plus A-ADT compared with those who received NA-ADT alone among the HR PCa cohort (HR: 0.46 [95%CI: 0.29–0.75], p = 0.002). However, the reported impact on OS among retrospective studies differed depending on the duration of ADT. When comparing those who received >6-month ADT with those who received no or ≤6-month ADT, the improvement of OS by adding long-term ADT (>6 months) was not observed.^{18, 23} An increased ADT duration was associated with a lower all-cause or PCa-specific mortality rate (≥12 vs. <12 months, or ≥24 vs. <24 months).^24-27 On the other hand, several studies reported that ≥24-month ADT led to no improvement of OS.^{28, 29}

Impact on biochemical progression

Regarding IR PCa, seven studies investigated the impact of adding 4–6-month NA-ADT on high-dose RT.^{21, 22, 30-34} Among them, one study reported the significant improvement of time to BF by adding NA-ADT³¹ and one study reported its tendency (p = 0.0937) (Table 2).²² In addition, the benefit of adding NA-ADT was observed in the subgroup aged <70 years old³² and an unfavorable IR population.³³ The remaining three studies reported similar BF control between RT alone and short-term (4–6 months) NA-ADT.^{21, 30, 34} Among these studies, one reported improved BF by adding longer ADT (4–12 months) compared with RT alone.³⁰ No studies reported an improved outcome with RT alone compared with NA-ADT. In addition, according to the phase 3 PMH 9907 trial, which compared EBRT plus androgen receptor blocker monotherapy for 5 months with EBRT alone predominantly among IR PCa patients (94.6% of study population), no significant difference was observed in the BF rate (p = 0.32).¹⁷ Regarding the duration of ADT, one RCT¹⁴ and two retrospective studies^{18, 31} reported no improvement of biochemical control by adding >4–6-month ADT while one retrospective study reported a tendency toward improvement of BF-free survival by adding 4–12-month ADT compared with shorter-term ADT (1–3 months) (p = 0.06).³⁰ According to a Japanese multi-institutional retrospective analysis of proton beam therapy,³⁵ although the benefit of adding ADT (with vs. without ADT) to improve BF-free survival was evident (HR: 0.49 [95%CI: 0.26–0.93], p = 0.029), this benefit disappeared when comparing those who received ≤6-month ADT with those administered >6-month ADT (HR: 1.00 [95%CI: 0.37–2.69], p = 0.999). In their subgroup analysis, no benefit was observed among those with a single IR factor (HR: 0.82 [95%CI: 0.33–2.02], p = 0.664) while the benefit of adding ADT was evident among patients with multiple IR factors. According to the phase 3 DART01/05 GICOR trial,¹⁴ which compared the outcomes of long-term (28 months) with short-term (4 months) ADT in combination with high-dose EBRT, no benefits of adding long- over short-term ADT were observed among the IR PCa population (HR: 1.82 [95%CI: 0.76–4.33], p = 0.174). Furthermore, according to a retrospective study of high-dose EBRT (75–78 Gy in conventional fractions), which divided the IR population into three risk groups, the benefit of adding short-term ADT (<6 months) over RT alone was observed only among patients with Gleason score (GS) 4 + 3 or T2c disease.³³ Another retrospective study reported that the benefit of adding ADT was observed only among patients aged ≤70 years old.³²

For HR PCa, firstly, among five studies that investigated the benefit of adding <12-month ADT over RT alone,^{24, 26, 27, 34, 35} one study, which evaluated the benefit of adding ADT to high-dose RT (HDR brachytherapy plus EBRT), reported that improvement of BF-free survival was noted in D'Amico HR patients (p < 0.001).³⁴ Regarding the benefit of adding longer ADT over RT alone, one study, which compared EBRT plus longer ADT (median duration: 21 months) with EBRT alone, reported that BF-free survival was significantly better in the EBRT plus ADT cohort (p = 0.0319).²² Secondly, among studies investigating the impact of the ADT duration on RT, 4–6 months, 1 year, and 2 years were used as cutoff values for the ADT duration. Among the three studies that used 4–6 months as a cutoff,^{18, 35, 36} two (67%) reported the improvement of BF-free survival by adding longer ADT.^{35, 36} Among five studies that used 1 year as a cutoff,^{24-27, 29} four (80%) reported an improvement by adding longer ADT.^24-27 Similarly, among four studies that used 2 years as a cutoff,^{25, 26, 29, 35} two (50%) reported an improvement by adding longer ADT.^{25, 26} According to a Japanese multi-institutional retrospective analysis,³⁵ the benefits of long-term ADT were observed among patients with multiple HR factors (T3a-4, prostate-specific antigen [PSA] >20 ng/mL, or GS ≥8) when using 21 months as the cutoff for long-term ADT; however, no difference was observed between ≥21 and 6–21 months ADT among patients with single HR factors. Furthermore, in a Japanese phase 3 study that compared medium-term ADT (14 months) under the policy of the early initiation of salvage-ADT (PSA >5.0 ng/mL) and long-term ADT (5 years) among patients with T3-4N0M0 PCa treated with definitive EBRT (72 Gy in 36 fractions), no difference in non-metastatic castration-resistant PCa-free survival was observed (HR: 1.132 [95%CI: 0.744–1.722], p = 0.5619).¹⁶ Similarly, in the phase 3 DART01/05 GICOR study, which compared long-term (28 months) vs. short-term (4 months) ADT, although there was the tendency of better BF-free survival in the long-term ADT arm, the difference was not statistically significant (HR: 1.91 [95%CI: 0.97–3.77], p = 0.054).¹⁴

Toxicities due to ADT

As ADT-related toxicities, cardiovascular events have been mainly investigated. According to the phase 3 DART01/05 GICOR study,⁵⁴ which compared 28-month ADT (28 months of goserelin acetate and 2 months of bicalutamide) vs. 4-month ADT (4 months of goserelin acetate and 2 months of bicalutamide), there was no significant difference in the incidence of mortality due to cardiovascular events between the long- and short-term ADT arms (2.8% for long-term ADT group vs. 1.7% for short-term ADT group, p = 0.470) although the cumulative incidence of cardiovascular events was significantly higher in the former compared with latter (17.6 vs. 7.2% at 5 years, respectively, p = 0.014) (Table 3). Similarly, according to a population-based study of cardiovascular mortality among patients who received definitive EBRT with or without ADT, the use of ADT did not increase cardiovascular mortality.³⁷ In addition, according to three studies^{16, 27, 38} that evaluated the impact of ADT on non-PCa or all-cause mortality, the use of ADT or long-term ADT did not significantly increase the incidence of non-PCa death among IR and HR PCa populations. Of note, among patients with at least a single risk factor of coronary artery disease, the use of ADT was associated with a significantly increased risk of all-cause mortality in men with low-risk PCa (HR: 1.36 [95%CI: 1.07–1.74], p = 0.01) but not in men with IR PCa (HR: 1.13 [95%CI: 0.96–1.35], p = 0.15) or HR PCa (HR: 0.86 [95%CI: 0.66–1.13], p = 0.28).³⁸

Regarding toxicities other than cardiovascular diseases, increased risks of hip and distal forearm fracture³⁹ or gynecomastia¹⁷ were suggested.

Decrease in quality of life due to ADT

According to a phase 3 study that evaluated the benefit of 5-month bicalutamide among patients with IR and HR PCa who received EBRT, no marked effect of the addition of bicalutamide on the QOL was observed.¹⁷ In addition, another prospective study, which compared RT plus ADT for 4–8 months with RT alone, reported that there was no significant difference in the incidence of genitourinary toxicities requiring alpha-blocker medication (Table 4).⁴⁰ However, the authors reported that the time needed to return to baseline urinary symptoms assessed by the health-related QOL score (UCLA Prostate Cancer Index) was significantly shorter in the RT plus ADT (4–8 months) cohort compared with RT-alone cohort in univariate analysis (UVA) (p = 0.017) although the difference disappeared in multivariable analysis (MVA).⁴⁰ On the other hand, according to a retrospective study among Japanese patients who received IMRT plus ADT, factors affecting the difference in the International Prostate Symptom Score (IPSS) between the baseline and at 24 months were age, IPSS classification, pretreatment GU medications, and positive biopsy rates (p = 0.023, <0.001, 0.044, and 0.028 in MVA, respectively), while the ADT duration had no impact.⁴¹ Similarly, according to a retrospective study that investigated predictive factors for urinary toxicities after LDR brachytherapy with or without ADT, the use of NA-ADT was not associated with the time to IPSS resolution.⁴²

Regarding sexual function-related QOL, according to a study using a multicenter observational prospective registry in the United States, the RT plus ADT group made significantly more complaints of reduced penile size than the RT-alone group irrespective of the ADT duration (p = 0.016).⁴³ In addition, according to a multicenter prospective study of QOL involving 1201 patients with non-metastatic PCa treated with EBRT or brachytherapy, in which the health-related QOL was measured using Expanded Prostate Cancer Index Composite-26, among patients who received EBRT, adding neoadjuvant ADT with a duration of ≥1 year significantly decreased the sexual and hormonal/vitality domain responses.⁴⁴ Specifically, they included questions regarding the ability to achieve an erection, quality of erections, frequency of erections, ability to reach orgasm, ability to function sexually, and lack of energy.

Impact of ADT on late toxicities due to radiotherapy

Among two phase 3 RCTs, one RCT, which evaluated the benefit of 5-month bicalutamide among patients with IR and HR PCa who received EBRT, reported that no significant differences in late genitourinary (GU) or gastrointestinal (GI) toxicities were observed between the RT plus bicalutamide monotherapy arm and RT-alone arm (p = 0.41 for late GU toxicities and p = 0.55 for late GI toxicities) (Table 5).¹⁷ Similarly, according to another phase 3 study, which compared 28-month ADT (28 months of goserelin acetate and 2 months of bicalutamide) vs. 4-month ADT (4 months of goserelin acetate and 2 months of bicalutamide), no significant differences in late GU or rectal toxicities were observed between the long- and short-term ADT arms (p = 0.912 for late GU toxicities and p = 0.345 for late rectal toxicities).⁵⁴ Regarding two retrospective studies, in one analysis, which included 3424 Japanese patients who received HDR brachytherapy plus EBRT, compared with RT alone, the use of neoadjuvant ADT alone was not significantly correlated with the decrease of ≥grade 2 late GU (p = 0.310) or GI (p = 0.560) toxicities, while the use of neoadjuvant plus adjuvant ADT was significantly correlated with decreased ≥grade 2 late GU (p = 0.0) or GI (p = 0.004) toxicities.²⁰ Conversely, in the other retrospective study, which included 1313 Japanese patients who received LDR brachytherapy with or without EBRT, ≥grade 2 GU toxicities were significantly higher in the RT plus neoadjuvant ADT group compared with RT-alone group (p = 0.001).⁴² The interaction between the prostate volume and use of neoadjuvant ADT was considered one potential explanation for the contradictory results.