No one likes Androgen Deprivation Therapy (ADT). Even in the short term, there are hot flashes to contend with, fatigue, and loss of libido. Longer term, there may be cardiovascular side effects, bone loss, fat gain, brain fog, low red blood cell count, blood sugar dysregulation, and mood swings. What are the potential risks and benefits of taking occassional vacations (called "intermittent androgen deprivation therapy"- iADT) from continuous androgen deprivation therapy - cADT?
For a discussion of all randomized trials of iADT vs. cADT, see this link. I will focus on a few of the most important trials that had homogeneous patient populations and had the longest follow-up. Patients must focus on the situation (called the "setting") they currently find themselves in - the advisability of using iADT differs with the setting.
1. Active Surveillance
Active Surveillance for low- and some favorable intermediate-risk patients requires no interventions. There have been several trials of short-term use of hormonal therapies - see this link under the subheading "I. Systemic Hormonal Therapies." All trials suffer from "lead time bias," which means that progression is not detected until after the hormonal therapy wears off. Nothing is gained: one only buys time onactive surveillance with time on the hormonal therapy.
2. Localized High Risk, Pelvic Lymph Node-only (N1), or Salvage Radiation
ADT in these situations must be continuous. For localized high-risk prostate cancer (PCa), ADT must continue for 12 months if brachy-boost therapy is given, and 26 months if external beam therapy is given (see this link). If there are cancerous pelvic lymph nodes detected by conventional imaging, 3 years of ADT and 2 years of abiraterone is standard-of-care (see this link). If detected on PET scan only, there is less data. If it is adjuvant ADT given with salvage radiation, 4 months to 2 years may be used (see this link)
Stopping and restarting ADT is like stopping and restarting antibiotics - it selects for the most resistant strains. The resistant strains are killed by sustained use of ADT and possibly more intensive hormone therapy. If unleashed by a vacation, they are not easily killed by simply adding more ADT. One can cause an aggressive, resistant type of PCa to predominate by using iADT in this situation.
3. Recurrent PCa
It is controversial whether to use ADT at all when PCa detected only by rising PSA after attempted curative treatment. The PRESTO randomized clinical trial showed that in patients with rapid PSA doubling time, a year of intensive hormone therapy (with apalutamide) could delay progression for a long while. The EMBARK randomized clinical trial showed that in similar patients, treatment with enzalutamide slowed castration resistance and the incidence of metastases.
Crook et al. reported the results of a Canadian randomized clinical trial (NCIC- CTG PR.7) among 1,386 men with PSA recurrence after primary radiation therapy. They were randomized to iADT or cADT. The iADT protocol was:
- no one had detectable metastases on a bone scan/CT
- 8 months on, 8 months off
- Vacations continued unless PSA during the on cycle failed to fall below 4 ng/ml, or if the lowest PSA was 1 ng/ml above the nadir reached in the previous cycle, or if there was evidence of castration resistance (3 consecutive increases during an on cycle or clinical evidence).
After 7 years of follow-up:
- There were no significant differences in adverse events, except for hot flashes, libido, and urinary symptoms.
- Testosterone recovered to baseline in 35% in the iADT cohort.
- Only 29% of men who were potent at baseline had a recovery of potency
- Median survival was 9.1 years for cADT vs 8.8 years for iADT (not statistically different)
- Median prostate cancer mortality was 18% for iADT and 15% for cADT (not statistically different)
- After adjustment to allow for time to get back on ADT and monitor PSA for 3 consecutive rises (Figure S5), eliminating ascertainment bias, there was no difference in time to castration resistance.
- Those who achieved lower nadir testosterone in the first year had lower PCa-specific mortality (see this link). Those with a higher nadir testosterone, developed castration resistance sooner and died sooner.
For recurrent men, iADT may give equal oncological results. Results may be improved, however, with more effective androgen deprivation.
4. Recurrent (Stage M0) or Locally Advanced (Stage N1) HSPC
Patients in these situations would rarely opt for treatment with ADT only, because they can potentially be cured with salvage radiation and adjuvant (limited term) ADT.
In the ICELAND trial, 701 recurrent (no distant metastases) or locally advanced (metastases only in pelvic lymph nodes - Stage N1) patients from 20 European countries were randomized to receive iADT or cADT. The protocol was:
- All patients had ADT with Lupron for 6 months at the start
- The off-cycle was stopped when PSA increased to 2.5 ng/ml
- The on-cycle was stopped when PSA fell to 1.0 ng/ml
- Castration resistance was declared if there were 3 consecutive PSA increases above 4.0 ng/ml during the on-cycle.
The trial was stopped 36 months after randomization. In that time:
- Only 11% became castration-resistant
- There was no difference in castration resistance between iADT and cADT
- Testosterone only rose to about half of the baseline level during the off-cycle
- Overall survival was high (85%), and there was no significant difference between groups
- There were no differences in side effects or quality of life between iADT and cADT
While the trial didn't run long enough for meaningful differences to emerge, the lack of a quality-of-life benefit for iADT may dissuade men in this situation from attempting iADT.
For newly detected patients with cancerous pelvic lymph nodes, a STAMPEDE trial has proved that continuous use of 3 years of ADT and 2 years of abiraterone improves survival and delays castration resistance over ADT alone.
These patients can potentially be cured with whole pelvic radiation and intensive hormone therapy. Such potentially curative treatment has been suggested by retrospective studies. Prospective randomized clinical trials (like this one and this one)will determine whether radiation with adjuvant hormonal therapy can provide a curative option to patients who are newly diagnosed with cancerous pelvic lymph nodes.
5. Metastatic Hormone Sensitive PCa (mHSPC)
Maha Hussain et al. reported the results of the SWOG-S9346 RCT. They randomized 1,535 men who were sensitive to hormone therapy.
- Half were randomly assigned to cADT, half to iADT
- Half were recurrent after primary prostatectomy or radiation; half had not been treated previously
- All responded to 7 months of ADT with a PSA<4 ng/ml
- In the iADT cohort, vacation was ended when PSA hit 20 ng/ml (10 ng/ml at doctor's discretion), and was triggered when PSA fell to 4 ng/ml. They were taken off the study if PSA didn't fall to 4 ng/ml while on ADT, or if PSA rose again before the first 3 months of starting a vacation.
- Half had extensive metastases (ribs, long bones, viscera); half had minimal metastases (axial skeleton and lymph nodes).
- Median PSA at baseline was 41 ng/ml
After a median follow-up of 10 years:
- Overall survival was 5.8 years for cADT vs. 5.1 years for iADT
- Mortality increased 10% due to cADT (range 23% increase to a 1% decrease with 90% confidence).
- Because the median increase did not reach the prespecified 20% hazard ratio for a clinically important difference, but the confidence interval included it, and a small percent (1%) lived longer with iADT, the study was considered statistically inconclusive. In other words, iADT may have been inferior, although they couldn't prove it definitively with the statistics obtained.
- The iADT cohort had better erectile function and better mental health at 3 months but not thereafter.
- There was no difference in adverse events.
- Castration-resistance started after the same amount of time in each group, after adjustment for allowing the iADT group to prove rising PSA while receiving treatment (Figure S5).
A secondary analysis
of SWOG-S9346 trial combined with Medicare codes focused on the long-term side effects of iADT vs cADT. After 10 years:
- Thrombotic and ischemic events (e.g., heart, lung, or brain blockages and strokes) were significantly worse for iADT (33%) than for cADT (24%)
- There were no significant differences in bone, endocrine, or cognitive events.
6. Metastatic Castration-Resistant PCa (mCRPC)
One of the causes of castration resistance is that the androgen receptor (AR) amplifies (upregulates) in each cancer cell. Because there are now so many copies of the AR, it takes very little androgen to activate cell replication. Some patients suppose that once they are castration-resistant they can cease using ADT. The opposite is true. They are more sensitive than ever to even the slightest amount of testosterone.
While iADT cannot be used, some researchers at Johns Hopkins hypothesized that rapid alteration between high testosterone and very low testosterone, called Bipolar Androgen Therapy, or BAT, may be able to retard AR upregulation. In clinical trials so far, BAT seems to slow enzalutamide resistance in some men (20%-30%) but accelerates progression in others (see this link). There has been some progress: those men with high AR activity seem to benefit. The test is so far only available at Johns Hopkins. There are already concerns about safety, so patients should not attempt BAT outside of a carefully monitored clinical trial.
II. No Difference in Time to Castration Resistance
The original hope of iADT was that it would slow progression to castration resistance. This was based on tests in mice conducted by Bruchovsky
. In clinical trials, this early hope did not pan out. All of the randomized clinical trials with homogeneous patient populations: SWOG-S9346
, NCIC-CTG PR.7
, and EC507
confirmed that the time to castration resistance was not different for iADT and cADT. Any reduction of evolutionary selection pressure with iADT is balanced by the reduction in sheer numbers with sustained cADT. This is not surprising, given that intensive hormonal treatment with ADT and Zytiga, Xtandi, or Erleada has been proven to delay castration resistance in men who are mHSPC.
III. Testosterone Recovery with iADT
The only reason to take a vacation from ADT is if testosterone recovers enough to provide relief from symptoms. Nabid et al. reported the testosterone recovery to normal testosterone based on how long they received ADT and whether their testosterone level was normal or below normal at baseline. Their data is from men curatively treated with ADT and radiation for intermediate and high-risk PCa. This table summarizes their findings:
Percent who recover to normal
Years to recover to normal T
among patients normal at baseline
among patients below normal at baseline
For 82% of men who have normal testosterone (T) at baseline and start with 6 months of ADT, it will take 1.5 years of vacation for their T to recover to normal levels. And almost half of men with below normal T at baseline will never obtain normal T levels. Older age and diabetes decrease the odds of T recovery.
A man who desires an ADT vacation should consider whether the vacation is likely to last long enough to give him a sufficient break from symptoms. Because men will be starting their vacation from castration at baseline, most men will never get the desired break.
Ong et al. similarly reported on testosterone recovery based on several clinical trials. A third of the men never enjoyed recovery to normal testosterone (>300 ng/dl or > 10.5 nmol/L). Older men (>65) were 20-33% less likely to recover normal testosterone.
Time (median) to recovery to normal T (300 ng/dl)
Percent who recover to normal T
among patients average at baseline
among patients normal at baseline
Dai et al. reported the time it took to recover to normal level (>250 ng/dl) of T after one treatment with 9 months of ADT. T at baseline determined T recovery time.
T Recovery (> 250 ng/dl) Time
Percent who recover to normal
Tunn et al. reported specifically on men on iADT. They found that after 6 months of ADT, T recovery to baseline occurred in 79% in the first vacation, taking a median of 100 days. Recovery diminished in the second vacation - 65% recovered in 115 days. They also found no difference in time to castration resistance between iADT and cADT.
Relogulix (Orgovyx) is a fast-acting oral medication that provides ADT. It is a better choice for iADT. After taking it or Lupron for almost a year, T rose to normal levels within 3 months with Orgovyx but barely above castration level with Lupron (see this link).
IV. Earlier Use of iADT
It has been hypothesized that iADT is more beneficial if used earlier. PET scans may be able to identify occult metastases while they are still small enough for occasional hormone therapy to keep them in check. PRESTO and similar trials of short-term intensive use of second-line hormone therapies may be able to provide a longer and safer break from side effects.
V. iADT Protocols
There are no established iADT protocols. It is always based on judgment, and should be arrived at after discussion with one's oncologist. Here are a few factors to consider.
1. PSA after first ADT cycle
An undetectable (<0.2 ng/ml) PSA nadir, whether achieved rapidly or slowly, on the first cycle of ADT is prognostic for successful cycles with iADT. If the lowest PSA is still detectable after several months of ADT, especially if a high nadir is achieved rapidly, cADT may be a better strategy (see this link).
2. Time on On-Cycle and Off-Cycle
- The Pr.7 trial used an arbitrary protocol of 8 months both on- and off-cycle.
- Others continue the on-cycle for at least a year.
- The time off-cycle may be different from the duration of the on-cycle.
- One can use an arbitrary time for the on-cycle and use a different factor, say PSA doubling time, for the off-cycle.
- One can let the on-cycle duration be variable: say, 8 months for the first on-cycle, 12 months for the second, 18 months for the third.
- The on-cycle duration variability can be based on the previous on-cycle PSA history (e.g., increasing nadirs), or off-cycle factors (e.g., higher PSA in last off-cycle).
3. Testosterone level achieved on the Off-Cycle
Since there is no other known purpose of iADT than to maintain quality of life, it may make sense to continue the off-cycle until normal testosterone (T) is maintained for some length of time. Some men will never achieve normal T levels (see above), so there is no point in risking iADT. It is important to measure T at baseline and during the off-cycle.
4. Maximum PSA target on Off-Cycle
The SWOG trial of men with metastases visible on a bone scan/CT used a maximum PSA of 20 ng/ml (10 ng/ml at doctor's discretion) as the signal that it was time to end the vacation. Trials that look at non-metastatic patient groups usually use a much lower PSA threshold for ending the vacation (typically 4 ng/ml). PET scans can detect metastases at lower PSAs (see this link), so the threshold may be set somewhere in between. The patient can decide what threshold he is comfortable with.
5. PSA Doubling Time (PSADT)
It is reasonable to use rapid PSA doubling time (PSADT) as the signal that it is time to end the ADT vacation. Typical threshold PSADTs may be 3, 6, or 9 months, depending on the patient's degree of comfort. Here is a PSADT calculator.
6. Symptoms/Radiographic Progression
PSA will usually provide an early signal of progression. However, the appearance of new metastases, enlarged metastases, bone pain may provide evidence of progression even before PSA has picked it up. The Bone Scan Index may be a useful metric.
7. Finasteride on Off-Cycle
Retrospective analysis suggested that finasteride might be helpful in extending the vacation time (see this link). However, when it was tried in the prospective clinical AVIAS trial, there was no benefit in extending the off cycle by adding dutasteride.
We are learning how to use genomic tools as an aid to decision-making. Recurrent, non-metastatic men with a high risk of developing metastases (Decipher high-risk) may wish to think twice about iADT, while those with low-risk Decipher scores may feel more comfortable opting for iADT. A recent new test developed by Veracyte (the company that makes Decipher) may one day be able to identify patients who are very responsive to hormone therapy and may be able to try iADT. Johns Hopkins has developed a test of Androgen Receptor activity that may prove useful in determining whether iADT with BAT makes sense.
9. Metastasis Directed Therapy (MDT) using SBRT
Men who play "whack-a-mole" with metastases as they crop up usually see their PSA drop. This is because serum PSA comes mostly from larger, detectable metastases that have developed their own blood supply. It is tempting to use iADT when serum PSA is negligible; however it may be risky to take long ADT vacations when PSA has been artificially tampered with. See "The Perils and Pitfalls of Treating PSA."
10. Adaptive iADT
Several schemes have been proposed for determining the optimal way of using iADT. The goal of these mathematical models is to delay castration resistance. They make adjustments to ADT use based on each person's response, and change as his tumors evolve. They can be thought of as tailoring ADT to suit the individual. So far, they have only been used clinically in small pilot trials or proposed pre-clinically. Here are a few examples:
A Phase 1b Adaptive Androgen Deprivation Therapy Trial in Metastatic Castration Sensitive Prostate Cancer
Evolution-based mathematical models significantly prolong response to abiraterone in metastatic castrate-resistant prostate cancer and identify strategies to further improve outcomes
Integrating evolutionary dynamics into treatment of metastatic castrate-resistant prostate cancer
Practically scheduling hormone therapy for prostate cancer using a mathematical model
Intermittent Androgen Suppression: Estimating Parameters for Individual Patients Based on Initial PSA Data in Response to Androgen Deprivation Therapy
VI. Cardiovascular (CV) Adverse Events (MACE)
The most life-threatening adverse events that can be attributed to ADT are cardiovascular: MACE=Major Cardiovascular Event (myocardial infarction or stroke).
In the SWOG-S9346 trial by Hussain et al., thrombotic and ischemic events (e.g., heart, lung, or brain blockages and strokes) were significantly worse for iADT (33%) than for cADT (24%). CV events are highest during the first 6 months of initiating ADT. It may be that the fluctuating T and estrogen levels inherent in iADT causes coagulation. Oral estrogen, which results in peaks and valleys in serum concentration, was known to cause blood clots and was eliminated as a therapy for that reason.
However, a meta-analysis of iADT trials found no difference in CV events compared to cADT, but there was a difference in deaths due to CV events. Those who used iADT were 15% more likely to die of MACE.
Orgovyx has been found to have a better CV risk profile than Lupron, although MACE incidence is not high for either drug. Incidence of MACE was 2.9% for Orgovyx vs. 6.2% for Lupron (see this link). If MACE occurred, it usually occurred within 6 months of starting ADT, and was significantly higher in men with a history of it (3.6% for Orgovyx vs. 17.8% for Lupron). GnRH antagonists (like Orgovyx or Firmagon) may release proteins that prevent plaques from forming in cardiac blood vessels (see this link), while GnRH agonists (like Lupron) may increase plaque instability (see this link and this link). The PRONOUNCE trial, which would have been definitive, failed to adequately recruit, and was ended early with no difference in MACE at 1 year between Firmagon and Lupron. A retrospective study suggested that Firmagon had lower MACE than Lupron.
VII. Managing Other ADT Side Effects
It may be possible to at least manage some of the most annoying ADT side effects without resorting to iADT. Here are interventions that can be used now or on the horizon for some of the common side effects of ADT. Each patient must determine for himself the balance between interventions and risk of the side effects.
The small randomized EXTEND trial at Duke last year reported that men who undertook a supervised program of aerobics and weight training before and during hormone therapy were able to significantly improve their fatigue scores and quality of life relative to men who did not take the program. Similar programs have been found to combat fatigue in small randomized trials in Australia (see this link and this one), and the UK,
There is a Phase 3 trial of Exercise and Methylphenidate (Ritalin) at MD Anderson with results expected by Nov. 2023.
Stopping smoking and restricting alcohol use may be beneficial.
Weight-bearing exercise may be contraindicated in men with multiple bone metastases due to risk of fracture. Cardiovascular morbidity may preclude aerobic exercise in some men. One should always check with one's doctors before beginning an exercise regime.
It should be understood the loss of desire for sex (libido) is not necessarily accompanied by a loss of erectile function, Men are often still able to have functional erections while on ADT if they use Trimix. Many men and their partners learn to value intimacy without intercourse. In some men, the loss of libido leads to a deterioration of relationships and a loss of one's masculine self-concept causing psycho-social problems.
3. Bone Loss
Long-term use of ADT leads to loss of bone mineral density (BMD). Resistance exercise and estrogen patches (see this link) can help maintain BMD. Unless there is fracture risk from multiple metastases, bone restorative agents (Zometa or Xgeva) should not be used before castration resistance because their most serious side effects are cumulative.
Excess calcium intake should be avoided due to risk of hastening progression (see this link and this one), cardiovascular side effects and kidney stones. There is no evidence that Vitamin D has any benefit unless one has a deficiency (see this link); in fact, high exogenous doses can pull calcium out of bones. Bloodwork at baseline and annually can determine if there is a deficiency.
Men using ADT should have a DXA scan at baseline, and annually thereafter. Here is a link for calculating FRAX score.
4. Lean Body Mass/ Muscle strength
Exercise has been found to help maintain lean body mass and avoid fat accumulation in a large number of small randomized trials (see this link), including this one. Transdermal estrogen is also known to preserve lean body mass. Amgen was developing a drug that may one day be used to inhibit myostatin in men on ADT (see this link and this one).
5. Metabolic Syndrome/Hyperlipidemia
Body Mass Index (BMI) and metabolic syndrome increases are well-known effects of ADT. They have been found to worsen prognosis. Again, exercise can help ameliorate it.
6. Hyperglycemia/ Insulin Resistance
Pioglitazone (Actos) an FDA-approved insulin sensitizer is being investigated as an adjunct to ADT in this clinical trial.
7. Hot Flashes
The following are good, albeit imperfect, solutions for hot flashes: aural acupuncture, estrogen patches, Progesterones (Megace or Provera), venlafaxine (Effexor), and oxybutynin.
8. Mental Effects/Mood Swings/Depression
Mindfulness, psychotherapy and antidepressants may help. Vortioxetine (Brintellix), an antidepressant, has been used in a pre-clinical study to reverse the ADT-induced cognitive impairment in rats.
ADT may cause anemia in some men due to the lack of erythrocyte stimulation that testosterone would normally accomplish (see this link). Erythropoietin (Procrit) may be able to reverse it if red blood cell levels get seriously low.
VIII. Shared Decision-Making
If you are feeling lost and confused after reading all of the above considerations about iADT, you are not alone. This is a consequential decision that should not be made lightly. It is a good idea to have a detailed conversation with your oncologist and significant support network.