Concept and Viability of Androgen Annihilation for Advanced Prostate Cancer

Abstract

There remains no standard of care for patients with a rising prostate-specific antigen (PSA) after radical prostatectomy or radiation therapy but who have no radiographic metastases, even though this is the second largest group of prostate cancer (CaP) patients in the United States. Androgen deprivation therapy (ADT) may cure some men with advanced CaP based on single institution series and a randomized clinical trial of immediate versus delayed ADT for men found to have pelvic lymph node metastasis at the time of radical prostatectomy. ADT may be more effective when initiated for minimal disease burden, which can be detected using PSA after radical prostatectomy or radiation therapy, and if more complete disruption of the androgen axis using newer agents decreases the chance that androgen-sensitive cells survive to adapt to a low androgen environment. Androgens may be “annihilated” sing simultaneously a luteinizing hormone releasing hormone (LHRH) antagonist or agonist to inhibit testicular production of testosterone, a cytochrome P45017A1 (CYP17A1) inhibitor to diminish metabolism of testosterone via the adrenal pathway and dihydrotestosterone (DHT) via the backdoor pathway, a 5α-reductase inhibitor to diminish testosterone reduction to DHT and backdoor metabolism of progesterone substrates to DHT, and a newer anti-androgen to compete better with DHT for the androgen receptor ligand-binding domain. Early initiation of androgen annihilation for induction as part of planned intermittent ADT should be safe, may reduce tumor burden below a threshold that allows eradication by the immune system, and may cure many men who have failed definitive local therapy.

The Clinical Challenge Posed by Advanced Prostate Cancer

Earlier detection of prostate cancer (CaP) due to increased awareness and use of prostate-specific antigen (PSA) has changed the presentation of CaP from mostly advanced to mostly localized. In spite of earlier detection and improved local therapy, approximately 30% of men fail potentially curative treatment and CaP remains the second most common cause of death in American men. In 2013, an estimated 29,720 men will succumb from CaP in the US.¹ PSA can be used to identify patients who have failed radiation or operation, who are likely to develop metastatic disease, and who have minimal tumor burden. There remains no standard of care for patients with rising PSA but who have no radiographic metastases, even though this is the second largest group of CaP patients in the US.

ADT for Advanced CaP

ADT has been the standard initial therapy for metastatic disease for more than 6 decades,² but ADT is considered palliative and ADT is associated with long-term cardiovascular and metabolic risks. Continuous use of ADT predisposes to weight gain, hypertension, hyperlipidemia, insulin resistance and glucose intolerance, metabolic syndrome, osteoporosis, cardiovascular disease, cerebrovascular disease and cognitive decline.^3–5 An alternative approach is to reserve ADT until the time of metastatic progression and/or symptomatic disease. A meta-analysis informed the American Society of Clinical Oncology Guideline that concluded immediate compared to deferred ADT decreased CaP-specific mortality but increased non-specific CaP-specific mortality, and thus had no effect on overall survival.⁶ CaP recurs during ADT due to continued transactivation of androgen receptor.⁷ Elevated levels of androgen receptor⁸ or molecular alterations in androgen receptor can increase response to low (castrate) levels of androgens. More recently, CaP has been found to synthesize testicular androgens,^{7, 9} probably from dehydroepiandrosterone (DHEA) and androstenedione (ASD),^10–12 weak androgens produced by the adrenal glands (adrenal androgen pathway), or cholesterol (cholesterol pathway)^{13, 14} (Figure 1). Intra-prostatic DHT without testosterone as a precursor may result from the backdoor pathway, especially when androgen metabolism pathways are altered by treatment.¹⁵ These new insights into the mechanisms of failure of ADT allow speculation that earlier and more complete attack upon the androgen axis may enhance extent and duration of response and perhaps even cure men with advanced CaP.

Figure 1.

Androgen metabolism pathways

Can ADT Cure CaP?

Early ADT has been studied, and its benefits proven, in 3 randomized clinical trials when used as neoadjuvant or adjuvant therapy for patients with high-risk localized disease. Immediate ADT improved survival and may have cured some men who were found to have pelvic lymph node metastases at the time of radical prostatectomy; 17 [36%] immediately treated vs 28 [55%] delayed ADT patients had died with 11.9 years median follow-up.¹⁶ Benefit also was demonstrated when neoadjuvant/concurrent/adjuvant ADT was used with radiation for patients with locally advanced or high-risk disease.^17–19 The ADT regimens among the radiation trials varied from 2 years to lifelong. The optimal duration of ADT needed to achieve a survival benefit is not known, and the benefits observed may derive from ADT improving the efficacy of local therapy, or ADT controlling micro-metastases.

An intermediate position is to use ADT to induce remission but then follow an intermittent ADT program to minimize side effects while still providing the benefits of ADT. A review of 19 published Phase 2 studies and interim results from 8 Phase 3 studies suggests that intermittent ADT reduces side effects by decreasing exposure to ADT while not adversely impacting survival.²⁰ A natural extension of this logic is to use intermittent ADT earlier in the disease, which allows an opportunity to determine if induction ADT induces remission that can be labeled cure when the PSA criterion for a second cycle of intermittent ADT is never reached.

Can More Complete ADT Cure CaP?

Serum PSA levels can be used to identify patients who have failed operation or radiation and have minimal tumor burden. No randomized trial has examined the benefits of administering ADT versus observing patients who have relapsed biochemically, but since ADT is the only available treatment to community oncologists and urologists, ADT is becoming the de facto standard treatment. Adding additional agents to enhance standard ADT has the potential to increase extent and duration of response to ADT and even cure some men with detectable PSA after operation or radiation for clinically localized CaP.

Metastatic or locally advanced CaP (Figure 2, left column) is treated with ADT, which is considered delayed ADT. The extent of response is determined by the relative distribution of CaP cells among 3 compartments. Androgen-dependent CaP cells undergo apoptosis that decreases tumor volume.² Androgen-sensitive CaP cells survive and remain static.^{21, 22} The androgen-sensitive cells could adapt to a castrate androgen microenvironment by amplifying,²³ hyper-sensitizing²⁴ or mutating²⁵ their androgen receptor to allow transactivation by weak adrenal androgens or castrate levels of testicular androgens. In addition, these androgen-sensitive cells could alter their androgen metabolism pathway to produce testicular androgens⁷ from weak adrenal androgens.¹⁰ Androgen-independent CaP cells not only survive ADT but continue to grow. The volume and growth rate of androgen-independent CaP cells and the rate of adaptation of androgen-sensitive CaP cells to castrate levels of testicular androgens and their subsequent growth rate determine the duration of response to ADT.

Figure 2.

CaP response may be greater when ADT is delivered earlier and more completely. The original tumor is debulked but not cured by local therapy. Left column shows tumor growth under observation until ADT is delivered for symptoms or some arbitrary PSA threshold. Right column shows immediate ADT that decreases tumor volume, but may do so less than immediate androgen annihilation, and hence fail to diminish tumor volume below the threshold for immunological cure. Fill and size of rectangle indicates androgen responsiveness and tumor volume, respectively. □ Androgen-dependent CaP; ▧ Androgen-sensitive CaP; or ■ Androgen-independent CaP.

Radical prostatectomy or radiation therapy may control CaP within the prostate or radiation field, respectively. CaP that remains after failed local therapy has a reduced tumor volume that is composed of a similar distribution of CaP cells among the androgen-dependent, androgen-sensitive and androgen-independent phenotypes (Figure 2, right column). Application of ADT would eliminate androgen-dependent CaP cells so tumor volume declines to the volume of the androgen-sensitive and androgen-independent CaP cells. Androgen-dependent CaP cells probably fail to survive castrate levels of circulating androgens; reductions beyond a testosterone threshold produced no further declines in tumor volumes in preclinical studies.²⁶ However, the effect of further reduction in circulating and tissue testicular androgens using new agents remains unclear.²⁷ A pre-clinical finding supports this concept; the androgen-sensitive Dunning H tumor could be cured only when ADT was combined with chemotherapy when treatment was initiated at low tumor volumes.²¹ Tumor volume after ADT may be reduced further if the androgen-sensitive CaP cell compartment is reduced. The androgen-sensitive CaP cell compartment could be reduced if circulating androgens were lowered further, if production of testicular androgens by intracrine metabolism was curtailed, and if more effective androgen receptor blockade caused some or all of these cells to undergo apoptosis before they adapted and grew in a castrate microenvironment.

Androgen annihilation (Figure 3) attempts to deprive CaP cells that are “hanging on” after standard ADT of the necessary mechanisms to survive and adapt to their new castrate microenvironment by intensifying the attack upon the androgen axis. Further attack upon the androgen axis may prevent or decrease the likelihood of androgen-sensitive CaP cells adapting to castrate levels of testicular androgens by impairing the changes in androgen metabolism necessary to produce testicular androgens from weak adrenal androgens, by either augmenting the adrenal (CYP17A1) or backdoor (5α-reductase and CYP17A1) pathways. Immediate application of androgen annihilation may reduce residual tumor volume to that of only the androgen-independent CaP cells. Hence, the timing of initiation of ADT as well as the extent of the response to ADT will impact the tumor volume that remains after failed local therapy and ADT. Cure could result if the remaining tumor volume composed of only androgen-independent CaP cells was below a threshold that could be dealt with immunologically.

Figure 3.

Androgen annihilation delivered by simultaneous use of LHRH agonist or antagonist (1), CYP17A1 inhibitor (2), 5α-reductase inhibitor (3) and anti-androgen (4)

Can the Immune System “Finish the Job,” especially When Tumor Volume is Low

ADT causes apoptosis of androgen-dependent CaP cells that should deliver intracellular antigens to the immune system. Theoretically, ADT may auto-vaccinate one against one’s own CaP. Pre-clinical data provide support for the concept of combining ADT and immunotherapy.^{28, 29} Neo-adjuvant ADT was associated with infiltration of the prostate by lymphocytes, macrophages and dendritic cells and inflammation.³⁰ The mechanisms whereby ADT can improve immune response by increasing T cells produced by the thymus, increasing B cells produced by the bone marrow, and decreasing T regs infiltrating tumor have been reviewed.³¹ The Immunotherapy for Prostate AdenoCarcinoma Treatment (IMPACT) study demonstrated that sipuleucel-T extended median survival by 4.1 months and improved 3 year survival by 38% compared to placebo³² when administered to men with castration-recurrent CaP. If immunomodulation produces survival benefit in advanced disease, immunomodulation (or the native immune response) may be more beneficial when minimal CaP remains after failed local therapy.

Cure for metastatic CaP has been described rarely, but examination of survival plots always reveals about 20% of men alive 4 years after ADT, although numbers at risk are small and extended follow-up is reported rarely.^33–36 A subgroup analysis suggested that patients with minimal tumor burden had better responses to complete androgen blockade than those treated with an LHRH agonist alone.³⁴ A secondary analysis of another randomized trial of men with locally advanced or asymptomatic metastatic disease showed early intervention with ADT was more effective than deferred intervention.³⁷ ADT improves survival and may cure some patients when applied in the adjuvant setting after definitive local therapy, especially when initiated when metastatic burden is low, such as in men who were found to have pelvic lymph node metastases at the time of radical prostatectomy.^{16, 38} 47% of men who received immediate and continuous ADT for lymph node metastatic disease were alive without evidence of disease compared to 14% of men in the deferred ADT arm. These results differ from population-based studies that demonstrated no benefit to early ADT³⁹ and are confounded by high rates of biochemical progression-free survival after operation alone.⁴⁰ A phase 3 trial of continuous versus intermittent ADT for men with PSA levels > 3 ng/ml after failure of primary or salvage radiation therapy⁴¹ produced similar disease-specific and overall survival and both arms demonstrated surprisingly low levels of CaP-specific mortality. After failed radical prostatectomy, a case series reported that all 52 men who underwent induction ADT achieved an undetectable serum PSA (PSA < 0.1 ng/ml) and that response continued in 9 men after median follow-up 62 months (range 20+ to 87+ months).⁴² These men who may be cured had usually suffered biochemical recurrence only after failed local therapy (n=5) or had clinically localized CaP (n=3). These findings have led some clinicians to suggest that more complete ADT for CaP should be initiated at the earliest time, when CaP tumor burden is smallest.^43–45

Evidence that More Complete ADT is Safe and Perhaps Efficacious

Androgen annihilation (Figure 3) could include simultaneous use of 1) an LHRH agonist (leuprolide or goserelin) or antagonist (degerelix); 2) an older large (bicalutamide or flutamide) or newer small (MDV3100 or ARN509) anti-androgen; 3) a non-specific CYP17A1 inhibitor (ketoconazole) or a newer, more specific CYP17A1 inhibitor (abiraterone acetate, TAK-700 or Tok-001); and a 5α-reductase inhibitor that inhibits 5α-reductase type 2 (finasteride) or 5α-reductase types 1 and 2 (dutasteride). Previous trials have demonstrated safety and sometimes efficacy of drugs proposed for androgen annihilation when used alone or in combinations of up to 3 drugs.^{42–44, 46–49} Tissue levels of testicular androgens and androgen-regulated gene expression persisted when LHRH agonist, bicalutamide, dutasteride and ketoconazole were administered for 3 months prior to radical prostatectomy.⁵⁰ Tumors were intermediate to high risk and probably bulky but, of 35 men, 2 had no residual cancer and 9 had tumor volume < 0.2 cm³. More complete ADT may remain non-curative, clonally simplify the tumor, and hasten growth and progression.^{51, 52} Well designed clinical trials are essential since a member of each of the 4 drug classes has not been used together in men with CaP.

LHRH agonist vs. antagonist

LHRH antagonists (abarelix and degarelix) are chemically modified LHRH, and hence pure antagonist peptides, that have pharmacological activity against CaP.⁵³ A Phase 2 study revealed that abarelix given as monotherapy did not induce an initial surge of testosterone, DHT, luteinizing hormone or follicle-stimulating hormone and that castration was almost immediate after the initial dose.^{53, 54} A multicenter Phase 3 trial compared the efficacy of LHRH agonist and anti-androgen against aberelix alone in patients with advanced CaP.⁵⁵ Aberelix achieved castrate serum level of testosterone more rapidly than combination therapy and avoided the testosterone surge characteristic of agonist therapy but the 2 treatments similarly reduced serum PSA, and achieved and maintained castrate levels of testosterone.

Anti-androgen

Bicalutamide and flutamide are “older” anti-androgens that are used with LHRH agonists in the first 7–10 days of treatment to prevent androgen flare in men with skeletal metastases in weight bearing bones or symptoms. However, usage as a single agent has been shown to be effective in the treatment of advanced CaP.^{43, 56} Nevertheless, there is a sense in the urological community, not founded in clinical trials, that flutamide alone is not sufficient for ADT because it causes a reflex increase in luteinizing hormone and testosterone levels. Furthermore, in vitro data suggest that flutamide, at high concentrations, can interact with the androgen receptor.⁵⁷ Finally, bicalutamide even at 150 mg/daily produced no improvement in overall survival compared to placebo.⁵⁸

MDV3100 (enzalutamide) is a second generation anti-androgen that showed selective, potent affinity for androgen receptor and lacked agonist androgen receptor activity in castration-recurrent CaP models.⁵⁹ Compared to bicalutamide, MDV3100 has a greater binding affinity to androgen receptor. MDV3100 extended survival by 5 months compared to placebo in patients with castration-recurrent CaP who were treated previously with docetaxel.⁶⁰ ARN-509 is an even newer anti-androgen⁶¹ about which less has been reported.

5α-reductase inhibitor

5α-reductase type 2 is the dominant 5α-reductase isozyme in benign prostate and 5α-reductase type 1 may be up-regulated in CaP; inhibition of both isozymes may be more effective for CaP treatment or prevention. Dutasteride, a dual 5α-reductase inhibitor, has been shown to decrease serum DHT to a greater extent than finasteride, a type 2 5α-reductase inhibitor (94.7% vs 70.8%).⁶² Treatment with dutasteride for 6–10 weeks before prostatectomy resulted in almost complete suppression of intraprostatic DHT, increased apoptosis and decreased microvessel density. A Phase 2 study of dutasteride for castration-recurrent CaP treated a total of 25 men with asymptomatic castration-recurrent CaP with 3.5 mg dutasteride daily for 2 to 9 months.⁶³ Dutasteride was safe but rarely produced even biochemical responses in men with castration-recurrent CaP.

LHRH agonist and anti-androgen

Combined androgen blockade (CAB) was reported initially to provide a substantial survival benefit in patients with metastatic CaP compared with either LHRH agonist or orchiectomy alone.^{34, 37, 64} Most now accept that CAB appears well tolerated but neither delays the time to progression nor increases duration of survival in patients with metastatic CaP. The largest, best designed trial randomized 1387 patients to orchiectomy with or without flutamide and showed no survival benefit for CAB.³⁵ Grade 2 or higher diarrhea and anemia were more frequent in the flutamide group than the placebo group (P = 0.002 and P = 0.024, respectively) although the incidence of toxicity associated with both treatments was low. When 603 patients with metastatic CaP were randomized to leuprolide plus placebo or flutamide, moderate diarrhea occurred more often in the flutamide group (P < 0.001) although tolerability was similar in both groups.³⁴ When 373 patients with locally advanced or metastatic CaP were randomized to goserelin or goserelin and flutamide, time to relief of bone pain was shorter in the CAB group, whereas the incidence of diarrhea and increased blood transaminases was higher in the CAB group.⁶⁵

Anti-androgen and 5α-reductase inhibitor

The combination of finasteride and flutamide in advanced CaP showed effectiveness and minimal side effects.⁴³ Twenty-two men with locally advanced (n=6) or lymph node metastatic (n=16) CaP were treated with finasteride and flutamide and 21 responded with decrease in PSA from mean 42.9 ng/ml to 3.6 ng/ml after 3 months and 3.1 ng/ml after 24 months. Sexual function was maintained in 86%, liver and renal function remained normal in all, 33% suffered diarrhea, and 19% developed gynecomastia; no patient discontinued treatment due to side effects.

LHRH agonist and anti-androgen and 5α-reductase inhibitor

Triple androgen blockade (LHRH agonist plus anti-androgen plus finasteride) was administered to 110 patients with clinical stage T1-T3 CaP for a median of 13 months followed by finasteride maintenance therapy.⁴⁷ All patients achieved undetectable PSA levels (<0.1 ng/ml), only 9 (8%) had PSA >4.0 ng/ml after median follow-up 3 years, all patients recovered normal testosterone levels, no patients had required reinstitution of treatment, and toxicity was minimal. Pubmed search revealed no update of this experience.

CYP17A1 inhibitors

Recognition of the importance of weak adrenal androgens as substrates for intracrine synthesis of testicular androgens was first tested by adrenalectomy for CaP recurrent after castration by Huggins,⁶⁶ then demonstrated by Geller,⁶⁷ confirmed by Mohler⁹ and then accepted by the field. Testicular androgens can be produced by recurrent CaP from adrenal androgens ^10–12 or cholesterol.^{13, 14}

CYP17A1 is an enzyme necessary for intracrine metabolism of testosterone from progesterone via the “cholesterol pathway”. CYP17A1 also is necessary to produce DHT from pregnen-3, 17-diol-20-one through the “backdoor pathway” (Figure 1). Ketoconazole is a non-specific weak inhibitor of CYP17A1 that has anti-tumor properties but considerable side effects. Testicular extracts and radio-labeled CYP17A1 steroid substrate identified abiraterone, a potent selective, irreversible inhibitor of CYP17A1. Continuous CYP17A1 inhibition increases levels of adrenocorticotropic hormone that can result in a syndrome of secondary minerocorticoid access. The minerocorticoid excess can be prevented by either a minerocorticoid receptor antagonist or low dose glucocorticoids. A Phase 3 study showed that median overall survival in patients randomized to abiraterone plus prednisone was 14.8 months vs. 10.9 months in those randomized to placebo plus prednisone (P < 0.001).⁴⁹ The study’s secondary endpoints also favored the treatment group, and were all statistically significant, including time to PSA progression (10.2 vs. 6.6 months), progression-free survival (5.6 vs 3.6 months), and PSA response rate (29% vs 6%). Adverse events were more commonly observed in the abiraterone group (comparing treatment with placebo for all grades) and included fluid retention (31% vs 22%) and hypokalemia (17% vs 8%). However, grade 3/4 hypokalemia (3.8% vs 0.8%) and grade 3/4 hypertension (1.3% vs 0.3%) were infrequent.

TAK-700 and Tok-001 (formerly VN/124-1) produced by Millennium/Takeda and Tokai Pharmaceuticals, respectively, are in clinical trials and each has theoretical advantages that may make them better drugs than abiraterone. TAK-700 is a napththylmethylimidazole derivative that selectively inhibits 17,20-lyse⁶⁸ but may offer an advantage over abiraterone since it is nonsteroidal and may not require coadministration of prednisone. TAK-700 has been characterized thoroughly in human cell lines and cynomolgus monkeys.⁶⁹ Two Phase 3 evaluations of TAK-700 400 mg twice daily continue in men with metastatic castration-recurrent CaP, which include the administration of concomitant prednisone; the post-docetaxel trial appears not to have met the primary endpoint.⁷⁰ 3-beta-hydroxy-17-(1H-benzimidazole-1-yl) androsta-5, or 16-diene (Tok-001) is another CYP17 inhibitor that has demonstrated anti-androgenic activity and is in preclinical development. Tok-001 is more potent than castration in a xenograft model.⁷¹ A unique feature of this compound is the ability to reduce AR protein levels, in contrast to bicalutamide or castration, which increased AR protein expression.⁷² Moving abiraterone and its relatives earlier in the course of ADT may produce responses of greater magnitude and duration and further the goal of turning CaP that recurs after failed curative therapy into a chronic disease.

Challenges

At each step in the processes of uptake, reservoir, reduction, degradation and excretion of testosterone and DHT, genetic polymorphisms in critical enzymes and proteins may result in substantial inter-individual variation that may affect the response to ADT or androgen annihilation. In addition, genetic polymorphisms may affect the pharmacokinetics of the different ADT agents that result in variation of effect that impacts the response of CaP. The causes of inter-patient variability in the extent and duration of response to ADT remain unknown, but could result in part from the variation in the level of AR activity observed among CaP,^{73, 74} polymorphisms in transporters that have been correlated with duration of response to ADT,^{75, 76} polymorphisms in 5α-reductase type 2 that have been associated with the risk of benign prostatic hyperplasia and CaP,⁷⁷ and responses to 5α-reductase inhibitors appear cell line-specific⁷⁸ (and probably patient-specific).⁷⁹ Finally, patients with recurrent CaP after radical prostatectomy and/or radiation therapy may suffer bowel and urinary symptoms and sexual dysfunction.^{80, 81} ADT may cause decreased energy, fatigue, depression and hot flashes.^82–84 The health-related quality of life (HRQOL) effect of androgen annihilation is unknown, and may be more prevalent and/or severe than that reported for standard ADT. If androgen annihilation improves duration of PSA response, HRQOL impact must be delineated to inform patient decision-making.