Is intermittent androgen-deprivation therapy beneficial for patients with advanced prostate cancer?



Use of intermittent androgen-deprivation therapy (IADT) in patients with prostate cancer has been evaluated in several studies, in an attempt to delay the development of castration resistance and reduce side-effects associated with ADT. However it is still not clear whether survival is adversely affected in patients treated with IADT. In this review, we explore the available data in an attempt to identify the most suitable candidate patients for IADT, and discuss factors that may inform appropriate patient stratification. ADT is first-line treatment for advanced/metastatic prostate cancer and is also recommended for use with definitive radiotherapy for high-risk localised prostate cancer. The changes in hormone levels induced by ADT can lead to short- and long-term side-effects which, although treatable in most cases, can significantly reduce the tolerability of ADT treatment. IADT has been investigated in several phase II and phase III studies in patients with locally advanced or metastatic prostate cancer, in an attempt to delay time to tumour progression and reduce the side-effect burden of ADT. In selected patient groups IADT is no less effective than continuous ADT, ameliorating the impact of ADT-related side-effects, and, to a degree, their impact on patient health-related quality of life (HRQL). Further comparative study is required, particularly in relation to HRQL and long-term complications associated with ADT.


Androgen-deprivation therapy (ADT) or androgen-suppression therapy has become the first-line treatment for advanced prostate cancer that has recurred or metastasised beyond the prostate and nearby tissues, into lymph nodes (LNs), bones, and other organs in the body [1, 2].

The changes in hormone levels induced by ADT can lead to short- and long-term side-effects including: reduced or absent libido (sexual desire); erectile dysfunction; hot flashes; gynaecomastia; osteoporosis; anaemia; decreased mental acuity; loss of muscle mass; weight gain; fatigue; depression; and increased risk of cardiovascular disease and diabetes [3]. Although treatable in most cases, these adverse sequelae can significantly reduce the tolerability of ADT treatment. In recent years, physicians have tended to alternate drug therapy with off-treatment periods in an effort to minimise side-effects without compromising outcome. Indeed, several phase II clinical studies have shown that intermittent ADT (IADT) can improve the tolerability of treatment [4].

Subsequently, several randomised phase III trials have suggested that IADT is associated with similar efficacy to that of continuous ADT (Table 1) [5-15]. The European Association of Urology (EAU) recently acknowledged that IADT should no longer be considered an experimental therapy [1]. However, not all patients appear to benefit in these studies [5], and despite the increasing number of clinical trials of IADT, it is still unclear which patients stand to benefit most. Here we explore the available data in an attempt to identify the most suitable candidate patients for IADT, and discuss factors that may inform appropriate patient stratification.

Table 1. Summary primary outcome and regimen design for reported IADT Phase III clinical trials
StudyPrimary outcomeDiagnosisRegimenInduction period, monthsPSA levels, ng/mL
Cease treatmentResume treatment
  1. CADT, continuous ADT; PFS, progression-free survival.
Calais da Silva et al. 2009 [5]No difference in survival between groups. No overall HRQL benefit was seen except for improved sexual activity in IADT group.Locally advanced or metastatic (M1 + M0); hormone naïve; PSA <100 ng/mL and ≥4 ng/mLGnRH agonist + cyproterone acetate, 200 mg/day3


or ≤80% baseline

≥10 (symptomatic) or ≥20 (asymptomatic) ≥20% above nadir PSA value
Calais da Silva et al. 2012 [6]No difference in OS. Improved HRQL (sexual activity) in IADT group.Advanced prostate cancerGnRH agonist + cyproterone acetate, 200 mg/day3NANA
Langenhuijsen et al. 2011 [7] TULPHigh baseline PSA, high nadir PSA and pain = poor prognosis. Patients with low PSA nadir do worse with IADT. No difference in HRQL.Advanced or locally advanced (metastatic) with baseline PSA ≥10 ng/mL (M0) or PSA ≥20 ng/mL (M1)Buserelin depot + nilutamide


Goal to achieve or maintain PSA <4 ng/mL and no clinical progression


≥10 (N0–3M0)

≥20 (N0–3M1)

Salonen et al. 2012 [8] FinnProstate Study VIINo difference in OS, prostate cancer-specific survival or time to treatment failure between groups.Locally advanced with PSA >20 ng/mL or metastatic with any PSA levelGoserelin, 3.6 mg every 28 days and 100 mg cyproterone acetate twice daily for first 12.5 days6<10.0 or at least 50% baseline> 20
Salonen et al. 2012 [8, 9] FinnProstate Study VIISignificant differences in HRQL were reported in activity limitation, physical capacity and sexual functioning in favour of IADTLocally advanced with PSA >20 ng/mLor metastatic with any PSA levelGoserelin, 3.6 mg every 28 days and 100 mg cyproterone acetate twice daily for first 12.5 days6<10.0 or at least 50% baseline> 20
Hussain et al. 2006 [10] Hussain et al. 2013 [11]IADT inferior to CADT. In patients with extensive disease IADT is non-inferior to CADTAdvanced (D2); hormone naïve; stratified by minimal and extensive diseaseGoserelin + bicalutamide7≤4
Crook et al. 2012 [12]Median OS was 8.8 years in IADT and 9.1 years in CADT groups. Some HRQL factors improved.Locally advanced (T1–4N1–3M0) or advanced (D1 or D2) no distant metastases*Goserelin + bicalutamide (4 weeks) at each cycle8


or ≥90% cf. baseline

Mottet et al. 2012 [13]No difference in OS or PFS survival between groups.Metastatic prostate cancer (D2) with bone metastases and PSA ≥20 ng/mLLeuprorelin 3.75 mg/month + flutamide 250 mg three times daily


Goal to achieve or maintain PSA <4 ng/mL and no clinical progression

<4≥10 or clinical symptomatic progression
Tunn et al. 2012 [14] (EC507)No difference in time to androgen-independent progression between groups.PSA relapse after RP; mostly M0/pT3/N0 patientsLeuprorelin acetate, 3-month depot + cyproterone acetate


Goal to achieve or maintain PSA <0.5 ng/mL and no clinical progression

<0.5>3 on two consecutive months
Calais da Silva et al. 2013 [15]No difference in OS between groups.Locally advanced or metastatic prostate cancer; cT3–cT4 M0 and M1, and PSA ≥4 ng/mL and ≤100 ng/mLTriptoreline 11.25 mg + cyproterone acetate, 200 mg/day


Goal to achieve PSA ≤4 ng/mL

NA≥20 or symptomatic

ADT Use in Patients with Advanced Prostate Cancer and the Rationale for IADT

ADT is first-line treatment for advanced/metastatic prostate cancer and is also recommended for use before, during, or after definitive radiotherapy for high-risk localised prostate cancer [1]. ADT may also be used for the provision of symptomatic control of prostate cancer in patients for whom definitive treatment with surgery or radiation is not possible or acceptable [16]. Many patients will, therefore, receive therapy over prolonged periods, and although the benefits of long-term treatment in these settings are well established [17], they are often accompanied by ADT-related side-effects. The side effects are a consequence of ADT-induced sex steroid deficiency and their impact on morbidity and health-related quality of life (HRQL) are variable [18-21].

The initial decision to instigate ADT will have considered both, the clinical benefits and potential short- and long-term side-effects of treatment, and ongoing patient management will incorporate measures to prevent and/or manage ADT-related complications. Recent evidence suggests the benefits of immediate instigation of ADT may be more evident in high-risk patients (PSA level of >50 ng/mL or a PSA doubling time of <12 months), whereas in patients with a more benign prognosis, deferred treatment is safe and significantly reduces the time on ADT, if indeed required at all [22].

However, increased awareness of the potential negative impact of ADT coupled with earlier diagnoses of prostate cancer since the introduction of PSA testing, support the concept of reducing patient exposure to ADT. The rationale for IADT is that alternating androgen blockade with treatment cessation allows hormonal recovery between treatment cycles, restoring the apoptotic potential and helping tumour cells to remain sensitive to treatment [4]. At the individual patient level, issues such as age, life-expectancy, and comorbidities will ultimately decide the appropriateness of IADT as an alternative to continuous therapy.

Best Candidate Patients for IADT

Many phase II studies of IADT have been performed over the last 15 years [4]. The designs of these studies have exhibited a significant degree of heterogeneity with regards to patients, PSA thresholds and cycle lengths. Treatment regimens generally included a GnRH agonist [4]. Results from several phase III studies have been published more recently, but none to date has reported a significant difference in overall survival (OS), although several have documented numerically superior OS for continuous therapy (Table 2) [5, 6, 8, 11-13, 15] (reviewed in [23-26]).

Table 2. Median OS reported in various phase III clinical trials
StudyMedian OS, monthsMedian follow-up, monthsHazard ratio (95% CI)P
  1. N, number of randomised patients for each treatment arm; CADT, continuous ADT; MAB, minimal androgen blockade; CAB, complete androgen blockade; *Pooled analysis two studies;**Median survival estimated using Kaplan–Meier method.
Crook et al. 2012 [12]105.6690109.269682.81.02 (0.86, 1.21)0.009
Calais da Silva et al. 2009 (MAB) [5]312314510.99 (0.80, 1.23)0.84
Calais da Silva et al. 2012 (MAB and CAB)* [6]7747690.98 (0.86, 1.12)0.77
Mottet et al. 2012 [13]42.28352.08644.40.75**
Salonen et al. 2012 [8] (FinnProstate Study VII)45.227445.7280651.15 (0.94, 1.40)0.17
Hussain et al. 2013 [11] (S9346)61.277079.6765117.61.10 (0.99, 1.23)>0.25
Calais da Silva et al. 2013 [15]462456660.90 (0.76, 1.07)0.99

Most randomised trials for IADT have comprised mixed patient populations of locally advanced and metastatic disease, making it difficult to clearly identify the best candidate patients for IADT. In addition, about a third of patients recruited into these trials does not achieve the required level of PSA to be placed on an IADT regimen and cannot, therefore, be randomised. Other factors, e.g. tumour grade (including presence of local and distant metastases), life-expectancy, comorbidities, patient preference, and sexual activity also warrant consideration when evaluating patients for IADT and when formulating patient inclusion criteria for trials of IADT.

Non-metastatic Patients

Patients with poorly differentiated prostate cancer and without clinically apparent bone metastases [27], as well as patients with local or biochemical failure after radiotherapy [12], have been suggested as likely candidates for IADT. Most of these patients will have no evidence of bone metastases.

In one phase III study, a better prognosis was documented in a subgroup of non-metastatic IADT patients aged <75 years, with a PSA level of <2 ng/mL and a Gleason score of <7. In these patients, disease progression was slower and HRQL better, suggesting patients with lower risk prostate cancer may be the best candidates for IADT [5, 6]. As the natural course of the disease in these patients is highly heterogeneous, factors such as PSA recurrence >3 years, PSADT of >12 months, no extraprostatic extension, seminal vesicle or LN involvement, may play an important role on the decision making of the added value of an ADT treatment.

High-risk disease is also highly heterogeneous [28]. However, adjuvant ADT has produced significant improvements in disease-free survival in men with high-risk features treated with definitive radiation [29, 30], and a significant OS benefit in men with Gleason scores ≥8 [29]. Patients with LN-positive (D1) disease after radical prostatectomy (RP) also have a survival benefit from adjuvant ADT [31], and a significant survival benefit was seen in a large study comparing lifelong ADT with delayed ADT in this group of patients [32].

In the Southwest Oncology Group (SWOG) 9921 study, 2 years of ADT after RP resulted in an extremely low rate of disease recurrence and prostate-cancer specific death for high-risk patients [33]. However, the optimal duration of ADT in high-risk localised disease is still under active investigation [34], and it has been suggested that many patients may not require such prolonged treatment, particularly those with indolent disease or a significant competing risk of death [23]. In a recent study of IADT compared with continuous ADT (which showed no difference in OS), the subgroup of patients with LN-positive disease exhibited a strong trend to improved time-to-progression with IADT (P = 0.07) [35].

Metastatic Patients

The advantages of IADT for patients with distant metastases (M1) are unclear and caution must be taken when deciding to place M1 patients on an IADT regimen. Ideally, these patients should have a relatively low tumour burden to be considered suitable for IADT. In one phase III study in newly diagnosed metastatic patients with a PSA level of >5 ng/mL (after a 7-month induction period), IADT was statistically non-inferior to continuous ADT and median survival was 7 months shorter in the patient group receiving IADT [11].

In another phase III randomised trial in patients with metastatic prostate cancer, no statistical difference was seen in OS or progression-free survival between the IADT and continuous ADT groups [13]. Although there were no significant differences in HRQL indicators between the groups, it was concluded that IADT appeared to be as safe as continuous ADT and could be useful in motivated patients, or those patients having significant treatment-related side-effects [13]. Similar results have been described in a previous study comprising a mixed patient population [35], suggesting that patients with high-grade prostate cancer that has spread to the LNs but associated with no other metastases, who are sexually active and compliant with ADT but intolerant of ADT adverse events, might benefit from an IADT regimen.

HRQL data from a study of IADT in patients with bone metastases revealed advantages in two out of five HRQL domains measured, although cognitive function appeared to be reduced in the IADT group [36]. Collectively, the results from these studies highlight the importance of careful patient selection and stratification when considering IADT in metastatic patients. It must also be recognised that the benefit of IADT to men with metastatic prostate cancer will inevitably be restricted by decreased life-expectancy and shortened off-treatment periods.

Poor Candidates for IADT

IADT is unlikely to be beneficial for patients with bulky tumours, multiple metastases in LNs and bone, high initial PSA levels (>100 ng/mL), severe pain, or those with extensive disease [37]. Also, patients who fail to achieve a PSA nadir of <4 ng/mL after 6 months induction [38] and those with a rapidly progressive PSA slope (>5 ng/mL/month) are poor candidates for IADT [37]. Two large trials reported recently, each gave insights into patient subgroups unlikely to benefit from IADT. In the first of these, patients with a Gleason score of 8–10 who received IADT had a trend towards worse OS [12]. In the second study, in patients with newly diagnosed prostate cancer with LN, visceral or bone metastases, IADT was also associated with a trend to worse OS, raising concerns that intermittent therapy may compromise survival in this patient population [11].


Serum PSA levels correlate with disease stage and recurrence [39], and it was recently shown that a PSA level of ≤4 ng/mL after 7 months ADT, is a strong predictor of survival in patients with newly metastatic prostate cancer (P < 0.001) [10]. In addition, patients with a PSA level of ≤0.2 ng/mL had a significantly better chance of survival than those with higher PSA levels [10]. Based on this and other randomised studies, IADT might be appropriate for those patients who achieve a PSA level of <4 ng/mL (or preferably <2 ng/mL) [10], after 6 months induction with a combined androgen blockade.

In a recent analysis of two studies that evaluated the combination of radiotherapy plus 6 months ADT in men with localised or locally advanced prostate cancer, PSA nadir and PSA post-treatment levels of >5 ng/mL were associated with an increased risk of prostate cancer-specific mortality [40]. For patients with biochemical recurrence after localised treatment, PSADT might also be useful to identify patients at high-risk of dying from the disease because, in general, shorter PSADTs are associated with more aggressive cancers [41]. In a retrospective analysis of data from IADT patients comprising a similar patient population, it was indeed shown that PSADT before treatment and during the first off-treatment interval was associated with disease progression [42]. Prospective validation of these data may help distinguish early in the course of treatment those who are more likely to benefit from IADT from those who may be candidates for other more aggressive approaches.

Testosterone Levels and IADT

Observed improvements in HRQL during IADT are a function of the rate of recovery of testosterone levels during the off-treatment period. These can be variable, and depend on the duration of ADT treatment, the patient's age, baseline testosterone levels, and ethnicity [43]. The off-treatment duration reported in prospective phase III trials of IADT ranged from 50% to 82% [5, 10, 14, 31, 35, 36], levels that are higher than those reported in phase II trials, in which patients spent a mean of 39% of their time off-treatment [44]. Most studies have also shown that the off-treatment duration diminishes with successive cycles (despite the slower recovery of testosterone). This presumably reflects the acquisition of a castration-resistant phenotype, with more rapid recovery of PSA [23].

Increasing testosterone to within the normal range can reduce the risk of cardiovascular disease and osteoporosis [45]; however, for sexual function, it has been suggested that the preservation of normal testosterone levels is not the only factor required to maintain this in elderly men [46]. Although testosterone levels can recover to normal after the first cycle of IADT, the extent of recovery diminishes in subsequent cycles, and the longer the patient remains on an IADT treatment regimen, the lower the recovery levels become, which effectively means that for some patients, especially the elderly, IADT is a temporary solution [47]. This observation was confirmed in a recent study, in which testosterone levels were measured in 109 patients on IADT. Recovery of testosterone to baseline levels was achieved in 79.3% of patients in the first cycle and in 64.9% of patients during the second cycle [14].

The implication that testosterone levels increase during the off-treatment interval have no apparent adverse impact on survival and appears at odds with other recent data, which suggests that patient outcomes are superior on continuous ADT if testosterone levels are maintained below 20 ng/dL [48, 49]. This apparent contradiction may be explained by the heterogeneous sensitivity of prostate cancer cells to androgen exposure [23].

Although limited information exists on the absolute relationship between testosterone values and clinical outcomes, it is advisable to monitor testosterone levels at baseline, after the initial induction period and when a patient on IADT (or continuous ADT) exhibits a rising PSA level. Monitoring testosterone levels is also recommended in EAU guidelines in order to ensure that castrate levels (<50 ng/dL [10.735 nmol/L]) are maintained [1]. Follow-up of serum testosterone levels during the off-treatment interval is often not part of routine clinical practice; however, inclusion of testosterone measurements during IADT may help physicians tailor the extent of re-exposure to endogenous testosterone and androgen-based therapy to an individual patient's response to treatment. Furthermore, it has been shown that during IADT, the PSADT becomes shorter and is associated with testosterone recovery [42]. Therefore, monitoring both PSA and testosterone levels during therapy may help stratify patients, guide treatment decisions, and contribute to clinical trial design.

Evidence Synthesis

ADT has for many years been a mainstay of treatment for patients with advanced prostate cancer, and treatment approaches have included orchidectomy, oestrogens, LHRH agonists, antiandrogens, and more recently the GnRH antagonists [17]. However, although the benefits of long-term treatment are well-established, patients may have to live with various ADT-related side-effects for many years. IADT allows androgen deprivation to be alternated with off-treatment periods to facilitate periodic testosterone recovery, and has been investigated as an alternative strategy to continuous ADT, with the rationale that it may reduce the side-effect burden and improve patient HRQL without compromising efficacy. Theoretically, IADT also has the potential to delay progression to castration-resistant disease.

Current data from randomised clinical trials comparing IADT and continuous ADT do not show a significant difference between treatment regimens. For OS, IADT has been similarly efficacious to continuous ADT, with the exception of one study in patients with metastatic disease [37]. Data from recent phase III clinical trials also suggest that specific subgroups of patients may be better suited for IADT than others, implying a need for careful patient stratification [5, 10, 12, 14, 27, 35, 36]. However, several phase III studies have reported shorter median survivals with IADT [12, 13, 37]. The differences, while not statistically significant, are large enough to suggest they may be clinically relevant, and the studies provide important new evidence to guide appropriate selection of ADT regimens [50].

Results from the study by Crook et al. [12] suggest that IADT should be considered for most men with non-metastatic disease, especially older men with a Gleason score of ≤7, associated comorbidities, poor tolerance to continuous ADT and slow PSA rises in the off-treatment period [50]. Conversely, results from the Hussain et al. [11] study suggest that IADT should be used with caution in men with metastatic disease, as less benefit may be derived due to shorter off-treatment time, with less potential for testosterone recovery, and a reduced life-expectancy [50]. Based on a thorough review of the available literature and interim results from ongoing trials, it is evident that those patients with locally advanced or relapsing disease (local or biochemical failure) are suitable candidates for IADT. These patients are likely to have a better chance of survival than those with more advanced and metastatic disease, and in such cases any prospective HRQL improvements might be of importance. In particular, patients newly diagnosed with non-metastatic prostate cancer, who achieve a PSA level of <4 ng/mL (or preferably <2 ng/mL) [10] after 6 months induction treatment, are likely to be good candidates for IADT.

In most studies, HRQL is improved during the off-treatment period [5, 12, 14, 35, 50], although some studies have failed to show a HRQL benefit [5, 13]. The range of questionnaires used to evaluate HRQL may partly explain the observed differences in benefit, as may the variety of treatment methods that include a varied duration of treatment cycles, monthly or 3-month depot formulations, and use of combined treatments or monotherapy. However, IADT has been shown to be better tolerated than continuous therapy [14, 15, 27, 35, 36], and several studies have documented improvements in sexual function [5, 6, 12, 15, 36, 47, 51]. It is nevertheless clear that the benefits of IADT need further careful evaluation in prospective, randomised trials, to better inform treatment decisions.

The decision to start or to switch to IADT requires a systematic evaluation and a detailed discussion of relevant information with the individual patient about his expectations, in the context of the specific benefits and risks of IADT. In practice, decisions are often founded on empirical experience gathered by the treating physician; however, it is clear that, where appropriate, the patient should be involved in the decision-making process [50]. Effective disease management should include counselling to promote general health and frequent monitoring of clinical parameters to achieve the best possible outcome.

Monitoring patients on IADT is important, to promote compliance and maximise treatment benefit. For that purpose, the standardisation of PSA and testosterone measurements is desirable, and would be anticipated to facilitate patient stratification and improve the reliability of data comparisons between studies. Another important consideration with any IADT regimen relates to the initiation and cessation of treatment. Although various thresholds have been applied in randomised trials, we can conclude that suspending treatment might be best at a PSA level of <4 ng/mL and that treatment should restart when the PSADT is <6 months (Fig. 1).

Figure 1.

Decision tree for continuous ADT or IADT applicable for metastatic or non-metastatic prostate cancer.


In conclusion, it can be argued that in selected patient groups IADT is no less effective than continuous ADT, ameliorating the impact of ADT-related side-effects, and, to a degree, their impact on patient HRQL. It remains to be clarified if effective ADT of any type should reduce the serum testosterone level to <20 ng/dL. Although IADT is an appropriate option for many patients needing ADT for advanced or recurring prostate cancer, there are still insufficient data to determine if IADT has the potential to prevent or reverse the long-term complications associated with ADT.

In cases where IADT is considered, we think the following principles should be considered:

  1. Initial induction phase of hormone therapy should last 3–6 months.
  2. Initial induction phase should have a clear impact on PSA levels:
    1. For men with metastatic disease, the PSA level should fall to <4.0 ng/mL (with a PSA level of <0.2 ng/mL showing the best prognosis for the patient and preferably <2 ng/mL) [10].
    2. For men with recurrent disease the PSA level should fall to <0.5 ng/mL.
  3. Treatment should be re-initiated according to current EAU guidelines [1] when clinical progression of disease occurs, or when:
    1. PSA levels rise to between 10 and 15 ng/mL in men with metastatic disease.
    2. PSA levels rise to between 4 and 10 ng/mL in men with non-metastatic disease.
  4. When treatment is reinitiated, it should be continued for ≥ 6 months.
  5. PSA tests and clinical examinations should be conducted every 3 or 6 months.

The range of PSA levels recommended for stopping and re-initiating IADT reflect the continuing nature of the debate in this area.

Finally, IADT will not be effective for all patients, and may be of less value to men who have limited adverse effects on ADT. As we are not able to predict with certainty which patients will be more likely to respond to IADT, a careful discussion with each patient about the comparative risks and benefits of IADT vs continuous ADT is mandatory before implementation of an IADT regimen.


The authors acknowledge financial support from Ipsen Biopharm Limited for the Advisory Board meeting before the initiation of the manuscript. The authors were fully responsible for the content of this manuscript and editorial decisions concerning it.

Conflicts of Interest

J.I. reports personal fees from IPSEN Pharma, during the conduct of the study; personal fees from ASTELLAS Pharma, personal fees from SANOFI Pharma, personal fees from FERRING Pharma, outside the submitted work.

All other authors have no disclosures.


intermittent androgen-deprivation therapy


European Association of Urology


health-related quality of life


lymph node


overall survival


PSA doubling time


radical prostatectomy