Listening to the Biology of Prostate Cancer

Quick Search


By Stephen B. Strum, M.D.

Reprinted from PCRI Insights October 2001 vol. 4, no. 4

Hormonal manipulation plays an important part in the management of prostate cancer at all stages of this disease. The common denominator of these endocrine therapies is the alteration in biologic milieu characterized by androgen deprivation. At the most fundamental level, this means lowering of the blood (serum) testosterone level. At a more sophisticated level it means depriving tumor cells that are dependent on male hormone (androgen) from access to such substances. This interaction occurs between the tumor cell’s docking sites or “receptors” and the active chemical messengers i.e. the androgens.

Simply stated, we need to deprive the tumor cell of androgens that act as growth promoters. This is the reason that the term Androgen Deprivation Therapy or ADT more adequately describes our therapeutic methodology.

ADT, optimally administered, necessitates an in-depth consideration of multiple issues critical to the success of this therapeutic strategy. These are:

I. Modalities of Treatment to Deprive the PC Cell of Androgen

II. Evaluation of the Degree of Androgen Deprivation

III. Assessing the Tumor Cell Population(s) Targeted by ADT

A. Obtaining a Baseline Panel of Tumor Markers

B. Evaluating the Extent of Biomarker Response to ADT

C. Evaluating the Depth of Biomarker Response to ADT


I. Modalities of Treatment to Deprive the PC Cell of Androgen

Androgen Deprivation Therapy connotes form and function. It involves manipulating the hormonal environment of the man with prostate cancer (PC) by surgical means (orchiectomy) or by medical means (LH-RH agonists, LH-RH antagonists, anti-androgens, estrogenic compounds, 5 alpha reductase inhibitors, cytochrome P450 inhibitors, selective adrenal androgen inhibitors and prolactin inhibitors). All of these therapies have in common a mechanism that deprives the tumor of a trophic substance – androgen, be it testosterone or its metabolite DHT. Optimal ADT almost certainly requires a combination of therapies. Details of such therapeutic approaches are shown in Table 1, (shown on a separate page). Figure 1 shows the endocrine axis in PC and where some selected treatments are operative.

Figure 1 - The Endocrine Axis in PC

The therapy the physician selects to deprive the tumor cell population of androgen may have consequential effects on the course of PC. For example, work by Sciarra et al has shown that 37% of men undergoing orchiectomy have a reflex increase in the production of the adrenal androgen precursor androstenedione.2 Androstenedione is metabolized within the prostate cell (both benign and malignant prostate cells) into testosterone (see Insights July 1999, pp 3-4 and October 2000, page 4). If the physician assumes that orchiectomy has resulted in a castrate testosterone (< 20ng/dl) and does not monitor the serum testosterone, almost 40% of these patients face a significant risk of disease progression. If progressive PC occurs, it would likely be assumed to be a reflection of androgen independent PC. In fact, it may be due to the reflex stimulation of the pituitary-adrenal axis due to the lack of testosterone — the production of androstenedione — and the subsequent conversion of this androgen precursor to testosterone within the prostate cell. The body tries to maintain balance or homeostasis in regard to testosterone and in doing so uses its backup systems.

It is important to point out that other authors have not confirmed the above findings. Oefelein et al studied 35 patients undergoing orchiectomy and noted that at a median time of 33 months after orchiectomy the median total testosterone was 15 ng/dl (range 10-30) with 32 of the 35 patients having testosterone levels of 20 ng/dl or less.3

There are, however, other examples of how the method of ADT can impact patient outcome if the physician is not familiar with the pituitary-testicular and pituitary-adrenal axes as well as the intra-prostatic synthesis of testosterone. Ketoconazole, an excellent suppressor of testicular and adrenal androgens, will over time result in high levels of LH production by the pituitary in an attempt to override the suppression of testicular androgens. Significant to this finding, LH receptors4 as well as FSH receptors on the PC cell5 may turn out to be significant factors in tumor escape with therapies such as ketoconazole. Therefore, in patients on ketoconazole, we advise ongoing use of LHRH agonists or the addition of other agents such as DES that turn off pituitary LH. Knowing the different modalities to deprive the tumor cell of androgen mandates an understanding of the endocrine mechanisms that the body and the PC may employ to bypass these blockades.


II. Evaluation of the Degree of Androgen Deprivation

Androgen Deprivation connotes efficacy of function. It is difficult, if not impossible, to fully understand the results of ADT in PC when the functional effects of androgen deprivation have not been evaluated. Checking to see if a castrate serum level of testosterone (T) has been attained and maintained is a mandatory assessment of ADT adequacy. However, physicians administering ADT rarely order this inexpensive laboratory test (serum testosterone). In the thousands of patients I have seen seeking second opinion, the frequency of finding results of serum testosterone levels in the medical record is only about 5%. Can you imagine assessing a hypertensive patient without a blood pressure or a diabetic patient without a blood sugar? Serum testosterone levels MUST be shown to become castrate using most forms of ADT and to remain there with interval testing until the physician feels assured that this highly important goal has been achieved. Assessment of androgen deprivation therefore mandates that testosterone levels be obtained monthly during ADT until two successive “castrate” values are achieved. Castrate levels are defined as <20 ng/dl (<0.69 nM/L). This would confirm that the hormonal environment has been satisfactorily altered by ADT and that androgen deprivation, at least based on the assessment of serum levels of testosterone, has been achieved.

Oefelein and Cornum have reported failure to achieve castrate levels of testosterone during LHRH agonist therapy e.g. Lupron® or Zoladex®, using the “3-month” (84 day) formulation. Five of 38 men or 13% of patients failed to achieve levels of testosterone less than 20 ng/dl. The authors suggested orchiectomy be considered in such patients.6 We have described a differential evaluation of such patients in the October 2000 issue of Insights that takes into account not only the possible failure of the LHRH agonist but also the overproduction of adrenal androgens. In other words, differential testing can be performed to determine if adequate LH suppression of the pituitary by the LHRH agonist has occurred. If this is confirmed by a serum LH level of less than 1.0, then the adrenal androgen precursors DHEA-S and androstenedione are evaluated to see if they are elevated or are high normal in value. If this is confirmed, then the cause of the non-castrate testosterone is due to excessive adrenal androgen production and either an anti-androgen, ketoconazole or steroids would be indicated choices of therapy — but not an orchiectomy. This approach is a more scientific way to evaluate a patient prior to consideration of an orchiectomy.

Current studies are also evaluating the depth of testosterone deprivation with a recent report by Kitahara et al suggesting improved clinical responses with further testosterone suppression below 20 ng/dl.7

It should be pointed out that serum testosterone measurement, per se, may not represent the most comprehensive biologic assessment in our efforts to manipulate the hormonal environment. Orchiectomy, for example, may reduce serum testosterone levels by as much as 95% and achieve castrate levels and yet leave from 25% to 40% of tissue DHT still available to stimulate tumor cell growth.8,9 At the very least, however, we should be measuring serum testosterone to insure that we have attained castrate levels (<20 ng/dl) during ADT.3

Baseline and follow-up measurements of DHT (a metabolite of testosterone) are secondary steps that we have initiated in addition to assessing the adrenal androgen precursors DHEA-S and androstenedione. The latter two hormones are metabolized to T. Such an assessment at baseline (and when indicated during the course of treatment of patients with ADT) enhances our understanding of elements relating to the success or “ failure” of such treatments. This is reviewed in detail in the October 2000 issue of Insights.


III. Assessing the Tumor Cell Population(s) Targeted by ADT

Importantly, an assessment of the effectiveness of ADT necessitates understanding the tumor cell population(s) targeted by ADT. The biologic manipulation and strategy of ADT is based on the assumption that we are dealing with androgen-dependent PC (ADPC) and not androgen-independent PC (AIPC). We cannot fault ADT for a lack of efficacy if we are treating patients who have a significant tumor cell population whose growth is independent of androgen, i.e. AIPC. It is unreasonable to expect that ADT will cause apoptosis or G1 arrest in cell populations of AIPC. Ironically, ADT is most frequently employed in patients with metastatic disease who most likely have a significant component of AIPC. The majority of patients with advanced PC having clinical stages ranging from T3c to D2 have tumor clones that have undergone mutation with de-differentiation to AIPC. How is it possible to evaluate the efficacy of ADT in clinical trials involving advanced PC? After all, we must surely be dealing with significant numbers of patients having heterogeneous cell populations containing significant proportions of AIPC that would not be expected to respond to a therapy directed primarily at androgen dependent PC cells.

A. Obtaining a Baseline Panel of Tumor Markers

Function, therefore, mandates that we assess the cell population(s) that is (are) targeted by ADT. We can do this by observing the effects of ADT on the function of the tumor cell population(s) that may be present. If there are tumor cell clones that are secreting cell products (biologic markers or biomarkers) into the blood stream and the levels of these products are not markedly decreased by ADT, we can presume that such clones are independent of ADT and therefore part of the AIPC cell population. Therefore, we need to know what the baseline levels of various biologic markers are prior to ADT, and to see if all abnormal marker elevations are at the very least normalized by ADT. If, however, this is not achieved despite a castrate level of T, then we must presume AIPC to be present and reevaluate our treatment strategy. The biomarkers we have employed are PSA, PAP, CEA, NSE and CGA. We know that elevations of CGA and NSE together are consistent with a variant of PC called small cell carcinoma that is most sensitive to chemotherapy. In addition, significant elevations of CEA in combination with any other marker(s) are most commonly a manifestation of AIPC.

B. Evaluating the Extent of Biomarker Response to ADT

A tumor cell population that is essentially androgen-dependent would be expected to be highly responsive to ADT with drops in PSA to undetectable levels. Conversely, a mixed or heterogeneous cell population containing ADPC and AIPC would show a drop that did not reach <0.05 and sustain this level. A clue to such heterogeneity in the PC cell population can also be found in heterogeneity of the biologic markers expressed by the various tumor cell components. Therefore, an assessment we believe that has been grossly ignored is the evaluation of other tumor cell biomarkers that reflect an AIPC population. The groundwork for this concept was reported by Steineck et al in 1996.10

They evaluated the response to chemotherapy in patients diagnosed as having AIPC. As shown in Figure 2, those patients with a decrease of 50% or greater of either PSA or PAP had an average survival of 18.9 months or 11.8 months, respectively. In contrast, patients having a drop of 50% or greater in both PSA and PAP had a survival of 29.8 months.

Figure 2. PSA and/or PAP decline on treatment vs. survival in AIPC patients

This study pointed out the importance to survival of having both of the tumor cell markers responding to a specific treatment. This indicates that non-response or poor response to treatment of one or more of the markers indicate resistance to the treatment being given. We can use this same concept in our evaluation of men undergoing ADT. Here, a lack of concordant tumor marker (biomarker) response to ADT would indicate the presence of a tumor cell population resistant to ADT, i.e. AIPC.11

C. Evaluating the Depth of Biomarker Response to ADT

The lowest level of PSA lowering (PSA nadir or PSAN) during ADT also yields significant information that provides clues to distinguish between the presence of ADPC vs. AIPC. In the setting of less advanced disease, when the tumor cell population is expected to be predominantly ADPC, it appears possible to assess the effects of apoptosis and/or G1 arrest during ADT. The depth of the PSA nadir achieved on ADT is an expression of tumor cell sensitivity to ADT that in turn reflects the nature of the PC tumor cell population, assuming a castrate level of testosterone during ADT has been reached.12,13 Using a hypersensitive PSA assay, we define a sensitive PC population as one in which a PSA nadir of <0.05 ng/ml is achieved and maintained for at least six months. In patients consecutively treated with two-drug ADT (ADT2) or three-drug ADT (ADT3) in a community practice of oncology directed exclusively toward PC, 90% of patients treated with either of these approaches achieve and maintain PSA nadirs of <0.05 ng/ml within the first four months of treatment. In those patients who have not reached or maintained PSA levels of <0.05 ng/ml, we uncover the cell clones that reflect AIPC that were present prior to the initiation of ADT. In other words, we are not inducing AIPC but uncovering it by killing or suppressing ADPC while AIPC goes untreated and progresses.

It is therefore logical to conclude that patients who have a profound drop in PSA to undetectable levels and who maintain such levels on ADT are those with predominant ADPC populations.12,13 In such a setting, ADT is of therapeutic value as well as prognostic value since it distinguishes or discriminates responding patients with a sensitive tumor cell population. In contrast, those patients with a suboptimal response or poor response to ADT are those with significant AIPC populations where ADT would not be expected to show benefit.14 The urologic literature contains numerous publications correlating:

  • Less PC progression and a higher percentage drop in PSA after one, three or six months of ADT11,12,15
  • An improved patient prognosis with greater depth of the PSA nadir during ADT11,13,14,16-19
  • A low PSA nadir during neoadjuvant ADT with the finding of organ-confined disease at RP.20-21

Some of the notable studies on this subject are summarized in Table 2.

Table 2. Studies Supporting the Value of PSA Nadir or Percent Drop in PSA vs Prognosis

The PSA nadir with two-drug ADT (ADT2) has been shown by Zelefsky et al to predict the outcome in patients receiving 3D conformal radiation therapy.22 Patients who did not achieve a PSA nadir of 0.5 ng/ml or less after three months of ADT2 (LHRH-agonist + anti-androgen) had a 5-year relapse-free survival of 40% vs. a 70% 5-year relapse free survival if a nadir of 0.5 or less was achieved (see Table 3).

Table 3. PSA After 3 Months of ADT2 Predicts the Response To 3DCRT

Such a difference in outcome appears to relate to the presence or absence of a tumor cell population that is not responsive to ADT2 and is likely to have spread outside the RT ports by virtue of its AIPC nature.


Past Clinical Trials Involving ADT in Advanced PC

In our assessment of patients with T3c, D1 and D2 disease, we should assume that the presence of AIPC is a high risk. The same would be true for patients with Gleason scores of (4,3) or higher or DNA aneuploidy. If our assumptions are valid, we would not expect ADT to eradicate or markedly decrease a tumor cell population (or its cell products – the biomarkers) whose growth is independent of male hormone. We can use the elevation of other biologic markers as circumstantial evidence that AIPC is likely to be present, especially if ADT does not reduce these values to normal. This is also confirmed by a short-term response insofar as the decline of PSA on ADT. Therefore, attempts to review the studies comparing monotherapy (ADT1) against two drug therapy (ADT2) are extremely difficult since all of the issues discussed above have not been evaluated in these studies. There is no way of knowing if patients with ADPC or AIPC have been equally randomized in these studies since:

(1) Castrate testosterone levels were not confirmed in such studies
(2) PSA nadir was not assessed using a hypersensitive PSA
(3) The measurement of other biomarkers and their response to ADT that would be reflective of the presence of AIPC was never done.

A recent landmark study by Bolla et al demonstrated the value of RT in combination with three years of ADT in the setting of advanced local disease (T3-4) or high grade Gleason scores (8-10).23 The study did not stratify patients based on the initial response to ADT. Based on many of the publications cited here, it would be expected that those patients not having a profound drop in PSA and/or other markers in response to ADT would be those patients having AIPC and most likely distant disease. Those patients having a profound drop in PSA while on ADT would likely represent those who would have done well on ADT alone. Unfortunately, the patients were not evaluated in this light.

Another landmark paper by Messing et al evaluated survival in men with D1 disease at RP who were randomized to receive immediate ADT vs. ADT upon recurrence.24 The findings supported the hypothesis that those men who received ADT immediately after RP had a better survival and a reduced risk of recurrent PC. However, one could postulate that those men receiving ADT upon recurrence of disease most likely had bulkier disease with a greater chance of mutation to AIPC. If such patients treated with ADT had been evaluated as suggested here, it is probable that AIPC could have been identified earlier and therapies changed to an AIPC regimen with a greater chance of survival. Scholz et al did this in a recent pilot trial using high dose ketoconazole in men showing a rising PSA on conventional ADT.25



A significant response and improvement in survival resulting from ADT would be expected to be seen in patients when:

1) Exquisite sensitivity to ADT is demonstrated by a decrease in PSA to undetectable levels (<0.05 ng/ml) with the ability to maintain these levels over many (at least 6-12) months;

2) Castrate levels of testosterone (<20 ng/dl) have been achieved and maintained; and

3) No additional cell clones, as reflected by increases in PAP, CGA, NSE and/or CEA, are present that might reflect mutation to AIPC. This is especially true if elevations of any of these markers are not normalized by ADT or show evidence of a progressive rise.

The foregoing is a basic endocrine-based oncologic methodology to PC treatment. This approach needs to be discussed, debated and, assuming consensus, incorporated into our strategic approach to the treatment of prostate cancer. We must tailor the treatment to the nature of the disease. We must listen to the biology.

Ten Concepts You Should
Get Out of This Article

1. ADT involves any manipulation that prevents androgen from stimulating PC cell growth.

2. ADTs often result in reflex increases in trophic hormones e.g. LH, ACTH, which try to override ADT in order to restore testosterone (T) to normal levels. Hence, knowledge of the hormonal axes and the repercussions of androgen blockade are essential for therapeutic success.

3. Confirming that serum T is below 20 ng/dl (0.69 nM/L) is an essential step to optimizing ADT involving orchiectomy, an LHRH agonist or antagonist, diethylstilbestrol, high-dose ketoconazole, and estrogenic compounds.

4. New studies suggest that further lowering of T below 20 ng/dl may result in improved clinical responses in the treatment of PC. Such findings are sufficiently important to merit additional clinical evaluation for validation.

5. Differential assessment of the hormonal axis in patients with a rising PSA on ADT is a logical way to rule out inadequate ADT and prevent the patient from being misdiagnosed as having Androgen- Independent PC (AIPC).

6. Unfortunately ADT is most commonly used in patients with advanced PC who are highly likely to have a significant component of AIPC. In such patients, an inadequate response to ADT is not a failing of ADT but a problem in defining the nature of the tumor cell population.

7. Baseline marker evaluation beyond PSA i.e. PAP, CEA, NSE, CGA is important especially in patients at high risk for AIPC. We cannot rely solely on PSA as the only response indicator in the absence of other tumor markers that have a significant chance of elevation in high-risk patients.

8. The degree of drop of the baseline PSA and/or of any other abnormally elevated biomarkers in response to ADT can be considered a “biologic stress test” which allows an appraisal of the Androgen- Dependent PC vs. AIPC tumor cell population.

9. A heterogeneous elevation of biomarkers, at baseline or during the course of the disease, is a high-risk finding for tumor cell heterogeneity i.e. ADPC plus AIPC.

10. The hypersensitive or ultrasensitive PSA is an inexpensive way to indicate the sensitivity of the tumor cell population to treatment with ADT and to establish an earlier diagnosis of AIPC if a PSA nadir of <0.05 ng/ml is not achieved and maintained.

return to Section I



1. Bartsch G, Rittmaster RS, Klocker H: Dihydrotestosterone and the concept of 5a-reductase inhibition in human benign prostatic hyperplasia. Eur Urol 37:367-380, 2000.

2. Sciarra F, Sorcini G, Di Silverio F, et al: Plasma testosterone and androstenedione after orchiectomy in prostatic adenocarcinoma. Clin Endocrinol 2:101-109, 1973.

3. Oefelein MG, Feng A Scolieri MJ et al: Reassessment of the definition of castrate levels of testosterone: implications for clinical decision making. Urology 56:1021-1024, 2000.

4. Halmos G, Arencibia JM, Schally AV, et al: High incidence of receptors for luteinizing hormone-releasing hormone (LHRH) and LHRH receptor gene expression in human prostate cancers. J Urol 163:623-9, 2000.

5. Ben-Josef E, Yang SY, Ji TH, et al: Hormone-refractory prostate cancer cells express functional follicle-stimulating hormone receptor (FSHR). J Urol 161:970-6, 1999.

6. Oefelein MG and Cornum R: Failure to achieve castrate levels of testosterone during luteinizing hormone releasing hormone agonist therapy: the case for monitoring serum testosterone and a treatment decision algorithm. J Urol 164:726-729, 2000.

7. Kitahara S, Yoshida K, Ishizaka K, et al: Stronger suppression of serum testosterone and FSH levels by a synthetic estrogen than by castration or an LH-RH agonist. Endocr J 44:527-32, 1997.

8. Labrie F, Belanger A, Dupont A, et al: Science behind total androgen blockade: from gene to combination therapy. Clin Invest Med 16:475-92, 1993.

9. Geller J, Albert J, Loza D, et al: DHT concentrations in human prostate cancer tissue. J Clin Endocrinol Metab 46:440-444, 1978.

10. Steineck G, Kelly WK, Mazumdar M, et al: Acid phosphatase: defining a role in androgen-independent prostate cancer. Urology 47:719-26, 1996.

11. Matzkin H, Eber P, Todd B, et al: Prognostic significance of changes in prostate-specific markers after endocrine treatment of stage D2 prostatic cancer. Cancer 70:2302-9, 1992.

12. Arai Y, Yoshiki T, Yoshida O: Prognostic significance of prostate specific antigen in endocrine treatment for prostatic cancer. J Urol 144:1415-9, 1990.

13. Miller JI, Ahmann FR, Drach GW, et al: The clinical usefulness of serum prostate specific antigen after hormonal therapy of metastatic prostate cancer. J Urol 147:956-61, 1992.

14. Pace CM, Lam PM, Roehrborn CG, et al: Nadir PSA level as a predictor of androgen independent progression. J Urol 163:181a, 2000.

15. Dijkman GA, Janknegt RA, De Reijke TM, et al: Long-term efficacy and safety of nilutamide plus castration in advanced prostate cancer, and the significance of early prostate specific antigen normalization. International Anandron Study Group. J Urol 158:160-3, 1997.

16. Zagars GK, Sands ME, Pollack A, et al: Early androgen ablation for stage D1 (N1 to N3, M0) prostate cancer: prognostic variables and outcome. J Urol 151:1330-3, 1994.

17. Fowler JE Jr, Pandey P, Seaver LE, et al: Prostate specific antigen regression and progression after androgen deprivation for localized and metastatic prostate cancer. J Urol 153:1860-5, 1995.

18. Stamey TA, Kabalin JN, Ferrari M, et al: Prostate specific antigen in the diagnosis and treatment of adenocarcinoma of the prostate. IV. Antiandrogen treated patients. J Urol 141:1088-90, 1989.

19. Furuya Y, Akakura K, Akimoto S, et al: Pattern of progression and survival in hormonally treated metastatic prostate cancer Int J Urol 6:240-4, 1999.

20. Wildschutz Th, Janssen Th, Schulman C: Valeur prédictive du taux sérique de PSA chez des patients ayant bénéficié d’un traitment hormonal néoadjuvant avant prostatectomie radicale. Acta Urologica Belgica 63:79-82, 1995.

21. Hachiya T, Kobayashi K, Ichinose T, et al: Impact of androgen deprivation prior to radical prostatectomy for T1, T2 prostate cancer on the likelihood of curative surgery. Nippon Hinyokika Gakkai Zasshi 88:936-44, 1997.

22. Zelefsky MJ, Lyass O, Fuks Z, Wolfe T, Burman C, et al: Predictors of improved outcome for patients with localized prostate cancer treated with neoadjuvant androgen ablation therapy and three-dimensional conformal radiotherapy. J Clin Oncol 16:3380-3385, 1998.

23. Bolla M, Gonzalez D, Warde P, et al: Improved survival in patients with locally advanced prostate cancer treated with radiotherapy and goserelin. N Engl J Med 337:295-300, 1997.

24. Messing EM, Manola J, Sarosdy M, et al: Immediate hormonal therapy compared with observation after radical prostatectomy and pelvic lymphadenectomy in men with node-positive prostate cancer. N Engl J Med 341:1781-1788, 1999.

25. Scholz M, Strum S, Mittelman P: High-Dose Ketoconazole (Keto) and hydrocortisone for hormone refractory prostate cancer (HRPC). Proc Amer Soc Clin Oncol 19:370a, 2000.