Hormone Therapy in Prostate Cancer

Stephen B. Strum M.D., Mark C. Scholz M.D., Glenn Tisman M.D.
March 20, 1997

Hormonal Pathways in Prostate Cancer

An understanding of the hormonal pathways involved in prostate cancer (PC) is important in the management of this disease. Knowledge of the biochemical steps that lead to testosterone and dihydrotestosterone synthesis may allow for therapeutic maneuvers that can result in more complete androgen blockade. The potential importance of progestins and their metabolism to androstenedione and from there to testosterone and/or DHT is also more evident if one becomes familiar with these hormonal pathways(diagram 1)

The hypothalamic-pituitary axis and its relationship to testicular and adrenal androgens is helpful in understanding sites of action of pharmacologic agents and their feedback loops(diagram 2). The testes and the adrenals are target tissues that produce androgens in response to hormonal stimulation arising from the control center in the hypothalamic-pituitary axis. The testicular androgens interact with the prostate cell directly and also via further enzymatic conversion that takes place within the prostate cell. The adrenal androgens are metabolized to testosterone within the prostate cell. The importance of the intra-prostatic synthesis of these androgens leads us to a more comprehensive appreciation of the role of adrenal androgens in the biology of the prostate cell.

The three pages that follow illustrate the above in diagram or table format. Diagrams 1, and 2 are graphic views of these pathways and Table 1 shows the hormonal effect of specific treatments, the site of hormone action, and reflex changes as a result of the hormonal effect.

Diagram 1. Metabolic Pathways of Hormones Involved in Prostate Cancer

Diagram 2. The Prostate Cell in Relationship to the Hypothalamic - Pituitary Axis

Table 1: Androgen deprivation therapies: site(s) of activity, hormonal effects and reflex increases

Hormone Therapy (HT) in the Treatment of D-1 and D-2 PC

Should androgen deprivation (AD) be reserved for advanced symptomatic disease or should AD be initiated earlier in the course of illness? As discussed above, PC is largely dependent on male hormones(androgens) for its growth. AD is a fundamental part of the treatment of PC. Most tumor populations of PC consistent of a large compartment of hormone-dependent cells.

Tumor burden, genetic mutation and clinical crises

There is literature that relates an increasing frequency of genetic mutation to an increasing amount of tumor. Therefore tumor burden or volume of tumor seems to be associated with changes in the genetic makeup of PC. This may be a part of the basis for the occurrence of androgen-independence or the so-called hormone refractory state. The development of androgen-independence does not seem to occur when patients with smaller tumor burdens are treated. Additional medical literature cites an increasing tendency toward aneuploidy (abnormal DNA) with time. All of these findings would support the concept of earlier treatment of PC rather than waiting until the disease is bulky and able to produce clinical symptoms. From a clinical standpoint of maintaining quality of life, treatment of PC should be an elective process & not evoked by a medical crisis such as bone pain, cord compression, or ureteral obstruction.

Importance of tumor burden on survival

Data from at least two studies confirm our concerns that PC should be treated earlier in the course of illness, not later. Labrie et al reported on the effect of tumor burden on efficacy of CHB (Lupron + Flutamide) in D-2 patients. There study showed a 60% survival at 8 years with CHB when the pretreatment bone scan showed 1- 5 bone lesions vs only a 20% survival at 8 years if more than 5 lesions on bone scan were found. If CHB was used when there were 1- 5 lesions there was a 4.4 year survival advantage compared to patients with 6-10 lesions versus a 6 year survival advantage compared to 11-40 lesions. ¹

The Intergroup #0036 data showed the effect of tumor burden on the efficacy of CHB vs monotherapy (Lupron or orchiectomy) in D-2 patients. The median survival was almost 20 months greater with CHB over that with monotherapy if minimal disease was seen on scan and the ECOG performance status was 0-2. The definitions of minimal and maximal disease and the ECOG scale is shown below. The actual median survivals were 61 months with combination therapy vs 42 months with monotherapy. The median progression-free survival was 48 months on the combination arm vs 19 months on monotherapy arm with a median follow-up period of greater than 5 years^2,3

Minimal disease = axial skeleton (spine) and/or pelvic bones or nodes
Maximal disease = above + appendicular skeleton (ribs, skull, long bones) or visceral disease (lungs, liver)

ECOG Performance Status
0= asymptomatic
1= restricted in strenuous activity but can do light work
2= up and about more than 50% of the time; unable to do work activities
3= bedridden more than half the time
4= completely disabled, totally confined to bed

(return to top)

Importance of symptoms on survival Labrie et al presented survival data in relation to clinical symptoms of disease. Median survivals in D-2 patients were 5.47 years in the group having minimal symptoms compared to 2.71 years in those with moderate symptoms and 2.1 years in patients with severe symptoms.¹ The definitions for different levels of symptoms is shown below:

General symptoms: anorexia, weight loss, nausea, fatigue, neurological signs, or pain

Minimal symptoms: The absence of general symptoms + an ECOG performance status of 0 - 1.

Moderate symptoms: one general symptom of moderate severity and/or pain of moderate intensity and an ECOG status of between 0-2.

Severe symptoms: 1 general symptom of severe or very severe intensity, plus pain and an ECOG status of 1-3.

Results of HT on Survival in patients with D2 and D1 disease

Comparisons of survival in patients with D1 and D2 disease treated by HT are difficult to evaluate due to different patient populations and different HT regimens. A sense of survival data reported in various studies is shown in Table 2 below.

Table 2. HT in D2 and D1 Prostate Cancer

(return to top)

Stage	# Pts	Rx	Median Survival (years)
			3	5	8	10
Labrie	Lupron + flutamide vs number of bone lesions on median survival
D2 11-40	50	CHB2	45	18	10
D2 6-10	45	CHB2	59	30	17
D2 1-5	105	CHB2	82	66	58
Zagars4	Effect of early orchiectomy on cause specific survival
	D1	111	O	93	85	57
Zincke5	Effect of early orchiectomy on cause specific survival
D1	294	O + RP	91	89	77	77
	86	RP	91	78	70	50
Zincke	Effect of ploidy status on cause specific survival
D1	69	D	96	87	83	72
	107	non-D	95	78	62	50
Zincke	Effect of EET and ploidy status on disease-free survival
D1	50	O+RP-D	100	100	100	100	100
	34	O+RP-non-D	85	68	62	62	0
	19	RP-D	90	70	62	42	0
	73	RP-non-D	62	18	10	10	no data

There are no long-term published studies on use of CHB in clinically localized PC. Most of the studies using CHB in earlier staged PC involve neoadjuvant studies. Some studies use CHB in early PC as part of an IHB (Intermittent hormone blockade) approach. Long term follow-up is not available in either of these two treatment groups.

Prevention of clinical and biochemical flare

What is flare?
We know that when an LHRH agonist is first started, it paradoxically causes a rise in the pituitary hormone LH. The LH rise stimulates the testicles to make testosterone during the first 5-12 days after initiation of the LHRH agonist. This increase in testosterone stimulates prostate cancer cell growth and is termed flare.

Why is flare prevention important?
In patients with advanced disease with subclinical spinal cord compression, flare can precipitate full cord compression and paralysis. If there is PC growing close to a nerve root then flare could result in pain in the distribution of that nerve. In patients with PC involving lymph nodes close to the ureters, flare could increase nodal disease and cause early compression of the ureter(s). Obstruction of both ureters could lead to kidney failure. Increasing disease in bony sites often leads to bone pain during times of flare.

What is clinical flare vs biochemical flare?
When this increase in tumor burden causes signs of bone pain, or compression of a nerve root or spinal cord compression, or blockage of ureter(s) this is called clinical flare. If the increase in tumor burden does not result in symptoms but does result in an increase in PSA alone, this is caused biochemical flare. We would like to avoid any stimulation of tumor cell growth whether or not it is associated with clinical symptoms. The purpose of Flutamide, Casodex, Nilutamide or any anti-androgen prior to Lupron or Zoladex should be to block all flare reactions. Other agents like Nizoral that turn off testosterone production from the testes or DES or Cyproterone acetate that decrease LH also have been used to prevent flare.

Can we prevent flare?
If we saturate the androgen receptors with Casodex or Flutamide we can prevent both clinical and biochemical flare. What we propose is to study not only LH and testosterone levels after starting Lupron but also to measure PSA levels. We propose that patients first beginning hormone blockade therapy use Casodex or Flutamide for one week before starting Lupron or Zoladex and allow us to draw blood for PSA, Testosterone and LH at least at baseline and for the first 7 days after Lupron or Zoladex is given. If there is no rise in PSA then there is no biochemical flare despite the elevation of LH and Testosterone that is initially seen after starting Lupron or Zoladex. If flare prevention is complete, the use of an anti-androgen prior to LHRH agonist therapy should deter any biological effect of increased testosterone until the binding of the LHRH agonist to the LHRH receptor results in decreased LH production.

What if initial anti-androgen therapy does not block flare?
If there is block in PSA production by pretreatment with anti-androgens prior to use of LHRH agonists, then biochemical flare can be eliminated. If this is not possible, then the use of agents that decrease LHRH such as DES could be considered for flare prevention or the use of agents to decrease testosterone production such as Nizoral could be considered. The combined use of either agent with an anti-androgen prior to the use of LHRH agonist therapy would have to be studied to determine the optimal dosing for biochemical flare prevention. The possible use of combined anti-androgen and 5 alpha reductase inhibitor prior to use of an LHRH agonist would also be a consideration since this combination would block DHT production and also prevent Testosterone and remaining DHT from binding to the androgen receptor. No one has studied these approaches in preventing flare.

(return to top)

Hormone Therapy in Prostate Cancer

Part 2

Neoadjuvant CHB In recent years numerous television shows have depicted our judicial system. All of us are now familiar with the principle of the Miranda.......... You have the right to remain silent, you have the right to an attorney..... Conceptually, the Miranda is frequently applied in medical practice but perhaps not enough. It would go something like this:

 

You have the right to know your diagnosis, ---

You have the right to understand principles of evaluation and treatment, ---

You have the right to be familiar with the pros and cons of available treatment options.

Too often patients are not read their medical Miranda (MM). We believe that patients with prostate cancer (PC) should have their MM. This MMPC or medical Miranda for prostate cancer patients should, at the very least, communicate the fundamental concepts involved in the evaluation and treatment of PC. We believe that your MMPC should involve your right to understand the following:

The concept of organ-confined disease
Is the PC likely to be confined to the prostate or is there a high risk that it is not? We know that systemic disease cannot be cured with local therapies such as: radical prostatectomy (RP), external beam radiation (EBRT), brachytherapy, or cryosurgery. Patients are entitled to know what group they belong to in regards to risk for systemic disease. Are you in a high, middle or low risk group? Using three pieces of information, the baseline PSA, Gleason’s score and clinical stage, we can relay to the patient their likelihood of having organ-confined disease, capsular penetration, seminal vesicle and lymph node involvement.

We have termed these tables, The Partin Tables, from the work of Alan Partin MD of the Johns Hopkins Medical Center.⁶ We consider the Partin Tables to be the major prognostic paradigm for the 1990’s.

Perhaps the addition of RT-PCR or Complexed PSA may enhance the value of these tables or surpass them; this should be evaluated. For now, we consider a discussion of the findings of the Partin Tables an absolute must in the physician’s initial or early discussion with the patient. In addition, in these days of concern regarding the cost of medical care, the Partin Tables will save millions of dollars in needless expense that involves inappropriate use of local therapy in the setting of overwhelming risk of systemic disease. Moreover the Tables are also able to indicate negligible risk for lymph node disease which might spare the patient the need for lymph node sampling and/or perhaps the need for pelvic CT or MRI. These studies would have a minimal chance of detecting such low-risk disease. Similarly, other reports have been published relating to risk-benefit and cost-savings in PC and relate to lymph node⁷ and bone involvement⁸. We have currently reviewed the work of other investigators that have employed this concept of combined modality staging as used originally by Partin. We have assimilated these reports into software programs that predict the likelihood of clinical and pathologic outcomes. These are reviewed separately on our homepage in a paper called: Predictive and Prognostic Information in the Counseling of Patients Newly Diagnosed with Prostate Cancer. These publications include the work of Narayan, Bluestein, Lerner, Kleer, Pisansky and D’Amico and their colleagues.

The concept of determination of extent of disease or stage
This is intimately related to the first item noted above of the MMPC. How has the diagnosis of PC been evaluated? Has there been a comprehensive history and physical examination and basic laboratory tests that may have implications for my ability to receive treatment? Have the determinations of Gleason’s score, clinical stage and PSA been made? Do I need a bone scan, endorectal MRI, pelvic MRI or pelvic CT? Is a determination of RT-PCR status for PSA relevant to treatment decisions in my case? We believe these are reasonable questions to ask your evaluating physician with the expectation of answers.

The concept of neoadjuvant hormone blockade (NHB)
The use of upfront combination hormone blockade (CHB) with an anti-androgen (such as Casodex, Flutamide or Nilutamide) in combination with an LHRH agonist (such as Lupron or Zoladex), has been shown to decrease the frequency of positive surgical margins at the time of RP. This has been reported in at least 5 published studies, 4 of them randomized. Patients proceeding directly to RP without NHB have had approximately a 39% chance of positive surgical margins while those receiving NHB with CHB have had an average of ~ 13% positive surgical margins.

Study Group	% Surgical margins without NHB (# of patients)	% Surgical margins with NHB (# of patients) with 3 mos CHB
Labrie et al.9 1993	38.5% (65)	13% (77)
Fair et al.10 1993	33% (72)	10% (69)
Solomon et al.11 1993	35.3% (119)	11.5% (156)
Fair et al.12 1995 update	36% (92)	11% (92)
Gleave et al.13 1995	Not randomized	5% (36)-- 8 mos CHB
Soloway et al.14 1995	47% (144)	17% (137)
Totals	39.2% (420)excluding Fair’s 1st group	13.5% (462)excluding Gleave

Thus, in 4 randomized studies from different institutions, the percentage of pathologically positive surgical margins (PPSM) in patients undergoing RP without NHB has been essentially identical. The benefit of 3 months of NHB with a reduction to approximately 13.5% from 39% of PPSM in all 4 studies attests to the validity of these findings. The reduction to 5% in one non-randomized study raises the issue of what is the optimal treatment time with CHB in a neoadjuvant setting.¹³

The use of the Partin Tables may also allow us to possibly bypass NHB in patients with favorable Gleason scores(2-4), low PSA readings(0-4) and clinical stages T1a to T2a. Predictions of organ-confined in such a setting is 85 to 100%. However, the prediction for capsular penetration of 22% in the T1c patients in the above group is still worrisome and would warrant a clinical trial to evaluate the need for NHB. The Partin Tables, if reviewed routinely for each patient being considered for local therapy with any current modality, would point out the risk for capsular penetration, seminal vesicle and lymph node involvement thus highlighting the need for NHB.

The long-term effects on survival with NHB are not available. The time in follow-up of patients receiving NHB has been too short for this determination. In Fair’s study, the NHB subgroup with pathologically organ-confined disease (no surgical margin +, seminal vesicle + or node + patients) have shown the identical biochemical PSA relapse rate as those patients having RP only with pathologically organ-confined disease.¹² Therefore 3 months of upfront CHB before RP appears to be having a biological effect on relapse. It appears to be not just masking disease as some opponents to CHB and NHB claim. In fact, in Walsh’s book The Prostate neoadjuvant therapy is criticized with the statement: hormone therapy is not a vacuum cleaner-it can’t suck the cancer cells back into the prostate once they’ve escaped.¹⁵ If this really were the case then no neoadjuvant therapy for cancer should work. The data on converting non-resectable lung cancer to resectable with neoadjuvant chemotherapy or significantly improving survival with neoadjuvant therapy for head and neck cancer, breast cancer or soft-tissue sarcoma testifies to the validity of the concept of neoadjuvant therapy, be it hormonal or non-hormonal.

The median follow-up time in the Fair et al study is 24 months. Further follow-up over the next 5 years will be critical in the assessment of the value of NHB on survival. However, since NHB does increase the number of men who are found to have surgical margin negative disease, then this increasing benefit should in and of itself justify the use of NHB. This latter observation seems to have been largely ignored by many critics of NHB.

Let's look at 3 populations of "virtual" PC patients each comprised of 100 patients eventually going to RP. The first 100 receive no CHB and based on the literature 39 men will have positive margins and 61 will have negative margins. Those 61 men will have no biochemical evidence of progression at a rate of, let's say, 80% at 4 years (this is purely for discussion's sake but in truth depends on the preoperative PSA,