Neoadjuvant Therapy for Prostate Cancer: An Oncologist’s Perspective

Abstract

With increasing use of prostate-specific antigen as a screening tool, diagnosis of prostate cancer has undergone a stage migration toward early-stage disease. Although this has increased the proportion of men who are candidates for definitive, potentially curative therapy, it has also made clear the limitations of our current standard of care. Specifically, despite adequate local therapy, a significant proportion of men go on to develop progressive disease. Neoadjuvant systemic therapy is one approach that continues to be studied as a way to maximize cure rates in the setting of early-stage disease. This article reviews the current data regarding neoadjuvant therapy, both hormonal and chemotherapy, and discusses which men are appropriate candidates for this option.

Prostate cancer, the most common cancer among men in the United States, is potentially curable in patients with localized disease. For these men, radical prostatectomy or radiotherapy might definitively eradicate all disease, resulting in overall survival rates of approximately 75%.¹ For men whose disease lies outside the prostate, however, the likelihood of cure is lower, and overall survival rates drop to 55% for those with regionally extensive disease and to 15% for those with metastatic disease. Fortunately, with the discovery of prostate-specific antigen (PSA) and with the advent of its widespread use as a screening tool, the pattern of prostate cancer diagnosis has changed markedly, leading to an increasing proportion of men diagnosed with early-stage, potentially curable disease, and a decrease in mortality. Although this trend is encouraging, there remains considerable room for improvement. Of those patients diagnosed in the early stages of prostate cancer, despite adequate local therapy, 15%–40% still go on to develop progressive, usually systemic disease.² These recurrences are likely due to micrometastatic foci present at the time of local therapy rather than to inadequate local techniques. Research has begun to focus, therefore, on identifying patients at high risk of developing recurrent disease and on optimizing therapy to increase the number of those men who are cured. Neoadjuvant therapy, specifically, has been studied as a means of achieving the above aims, theoretically through downstaging the disease and eradicating subclinical deposits at the earliest possible time.

Herein, we will review the current data regarding neoadjuvant therapy, both hormonal and cytotoxic, selection of the most appropriate patients for such therapy, and future directions.

The Rationale for Neoadjuvant Therapy

The aim of adjuvant therapy is to maximize cure rates for patients who have undergone definitive therapy for localized disease, theoretically by eliminating micrometastatic disease. Neoadjuvant therapy, on the other hand, not only provides possible early systemic treatment for any subclinical distant disease, but also aims to help improve local disease control and to increase the number of patients who may be eligible for definitive local therapy via downstaging. Neoadjuvant therapy also has the benefit of allowing measurement of tumor response to the therapy applied. Not only does this enable assessment of the efficacy of a particular treatment in an individual patient, possibly even providing some sense of prognosis, but it also allows rapid screening of promising therapies. For instance, once a new drug is identified, whether from preclinical programs or from recognition of its benefit in a different tumor type, the time from phase I investigation to definitive evaluation of efficacy with phase III trials is measured in years. Furthermore, phase I trials with new agents often are conducted in patients with significantly advanced, heavily treated disease—a population that may be quite different biologically from patients with early-stage disease. By testing agents in a neoadjuvant setting, we may get some sense of efficacy and insight into determinants of response and resistance in early-stage disease.

The neoadjuvant approach in prostate cancer can be hampered, however, by our limited ability to quantify clinically localized disease. Although digital rectal examination (DRE) and transrectal ultrasonography (TRUS) are the most widely used tools to assess local extent of disease, they have an element of subjectivity, and neither has been validated for assessment of response. Based on preliminary data, magnetic resonance spectroscopy might be more sensitive in this setting, because it yields information on tumor viability as well as anatomy,³ but this technique is still in its clinical infancy. Even PSA, the most commonly used surrogate marker of response in advanced disease, does not correlate well with the extent of residual disease.⁴ Instead, pathologic endpoints are the most objectively measurable endpoints of neoadjuvant therapy. Although complete response, or pathologic downstaging to pT0, is the optimal outcome, it is not clear that this should be the only meaningful endpoint for neoadjuvant studies. Experience with chemotherapy for other solid tumors indicates that despite very low clinically complete response rates in the setting of advanced disease, survival benefits are attained with these agents in the adjuvant setting. Although their clinical significance is unclear, alternative pathologic endpoints for patients who achieve less than a complete response may include number of patients with negative margins, lack of extraprostatic extension, and node-negative disease. For those patients who go on to radiation therapy, pathologic endpoints are obviously more difficult to assess, and may include posttherapy biopsies.

Overall, as with any tumor type, certain principles dictate whether combined-modality therapy is appropriate: 1) an adequate local therapy should already exist; 2) there should be sufficient risk of recurrence or progression in the population at hand to warrant combined therapy; and 3) candidate drugs for neoadjuvant therapy should be agents that are active against the disease. As noted above, with both an effective local therapy and a significant recurrence rate in high-risk populations, the first two requirements are clearly met for prostate cancer. As for the third, certainly hormonal agents have been the mainstay of systemic therapy for prostate cancer, with response rates of 80%–90% in those with advanced disease and median progression-free survival of 12–33 months.⁵ Cytotoxic agents, although historically understood as inactive in prostate cancer, are increasingly finding a place in the treatment of this disease, as will be discussed below.

Selection of Appropriate Patients for Neoadjuvant Therapy

Clearly, to assess the potential benefit of neoadjuvant therapy, the population at risk needs to be accurately defined. Because stage migration has increased the number of men diagnosed with early-stage disease, the ability to detect other high-risk features has improved and continues to evolve.

Standard Clinicopathologic Criteria

Although predicting overall cause-specific survival from clinical or pathologic parameters would be ideal, the protracted course of prostate cancer precludes direct correlation of this endpoint with prognostic variables. Instead, intermediate markers, such as the prediction of pathologic extent of disease, biochemical treatment outcome, and disease-free intervals, have been used as endpoints. The pretreatment prediction of patients at high risk of failure after local therapy is largely based on three well-established variables: clinical tumor stage, biopsy Gleason score, and PSA level at diagnosis. Each has been validated as having independent prognostic value in predicting pathologic stage or biochemical outcome.⁶ In one series, for instance, men with a preoperative PSA level of lt; 10 ng/mL had a 70%–80% likelihood of organ-confined disease, as opposed to 25% in men with a PSA level > 50 ng/mL.⁷ Similarly, in another series, the 10-year post-radiotherapy risk of developing distant metastases was 13% for Gleason grades 2–3, 34% for grades 4–6, 52% for grade 7, and 63% for grades 8–10.⁸

To optimize the accuracy of these three predictive variables, several groups have combined them into various user-friendly preoperative nomograms.^9–11 These have been validated in multi-institutional settings as able to predict either pathologic stage or biochemical recurrence in patients treated with definitive local therapy.^12,13 The most widely used model is the set of Partin tables, which has recently been updated.⁹ According to this nomogram, for example, a man with clinical stage T2a disease, PSA level of 3 ng/mL, and a Gleason score of 4 has an 85% likelihood of having organ-confined disease; a man with the same clinical T-stage but a PSA level > 10 ng/mL and a Gleason score of 8, however, has only a 52% chance of having organ-confined disease (Table 1). The predictive data contained within these models facilitates the stratification of patients into specific risk categories, a practice that can guide the selection of patients for additional therapy.¹⁴ Ten-year risk of biochemical failure in low-risk patients, as defined by clinical T-stage T1c-2a, PSA level ≤ 10 ng/mL, and Gleason score ≤ 6, is 17%; for intermediate-risk patients, with stage T2b, PSA level of 11–20 ng/mL, and Gleason score 7, 54%; and for high-risk, with stage T2c, PSA level ≥ 20 ng/mL, and Gleason score ≥ 8, 71%.

Table 1.

Example (Clinical Stage T2a) from the Partin Tables, a Prostate Cancer Staging Nomogram

PSA Range		Gleason Score
(ng/mL)	Pathologic Stage	2–4	5–6	3 + 4 = 7	4 + 3 = 7	8–10
0–2.5	Organ confined	91 (79–98)	81 (77–85)	64 (56–71)	53(43–63)	47 (35–59)
	Extraprostatic extension	9 (2–21)	17 (13–21)	29 (23–36)	40 (30–49)	42 (32–53)
	Seminal vesicle (+)	-	1 (0–2)	5 (1–9)	4 (1–9)	7 (2–16)
	Lymph node (+)	-	0 (0–1)	2 (0–5)	3 (0–8)	3 (0–9)
2.6–4.0	Organ confined	85 (69–96)	71 (66–75)	50 (43–57)	39 (30–48)	33 (24–44)
	Extraprostatic extension	15 (4–31)	27 (23–31)	41 (35–48)	52 (43–61)	53 (44–63)
	Seminal vesicle (+)	-	2 (1–3)	7 (3–12)	6 (2–12)	10 (4–18)
	Lymph node (+)	-	0 (0–1)	2 (0–4)	2 (0–6)	3 (0–8)
4.1–6.0	Organ confined	81 (63–95)	66 (62–70)	44 (39–50)	33 (25–41)	28 (20–37)
	Extraprostatic extension	19 (5–37)	32 (28–36)	46 (40–52)	56 (48–64)	58 (49–66)
	Seminal vesicle (+)	-	1 (1–2)	5 (3–8)	5 (2–8)	8 (4–13)
	Lymph node (+)	-	1 (0–2)	4 (2–7)	6 (3–11)	6 (2–12)
6.1–10.0	Organ confined	76 (56–94)	58 (54–61)	35 (30–40)	25 (19–32)	21 (15–28)
	Extraprostatic extension	24 (6–44)	37 (34–41)	49 (43–54)	58 (51–66)	57 (48–65)
	Seminal vesicle (+)	-	4 (3–5)	13 (9–18)	11 (6–17)	17 (11–26)
	Lymph node (+)	-	1 (0–2)	3 (2–6)	5 (2–8)	5 (2–10)
> 10.0	Organ confined	65 (43–89)	42 (38–46)	20 (17–24)	14 (10–18)	11 (7–15)
	Extraprostatic extension	35 (11–57)	47 (43–52)	49 (43–55)	55 (46–64)	52 (41–62)
	Seminal vesicle (+)	-	6 (4–8)	16 (11–22)	13 (7–20)	19 (12–29)
	Lymph node (+)	-	4 (3–7)	14 (9–21)	18 (10–27)	17 (9–29)

Data represent percent chance (range) of having organ-confined disease. PSA, prostate-specific antigen. Reproduced with permission from Parti from Partin et al.⁹

Another clinicopathologic variable that may further clarify risk, particularly for the heterogeneous intermediate-risk category, is the percentage of positive biopsies. This is felt to be a marker of tumor volume and has recently been validated as a variable with additional predictive value for those at intermediate risk.¹⁵

Novel Imaging Techniques

Because TRUS, even in combination with DRE, has suboptimal sensitivity, specificity, and predictive values,¹⁶ newer imaging modalities have been developed. Endorectal magnetic resonance imaging may provide a more accurate assessment of tumor volume and may predict biochemical outcome.¹⁷ This technique was found to increase the accuracy of detecting localized disease, from 32% to 61%. It is unclear, however, whether this provides a significant addition to the above described variables.

Neoadjuvant Hormonal Therapy

As noted above, hormonal therapy plays a major role in the treatment of advanced prostate cancer. Because of this, and considering the hormone-dependent growth of most prostate cancer cells, hormonal manipulation has been studied fairly extensively in the neoadjuvant setting. Interestingly, the results have been different in the setting of radiotherapy compared with that of radical prostatectomy.

Neoadjuvant Hormonal Therapy and Radical Prostatectomy

Initial investigations into neoadjuvant hormonal therapy focused heavily on the pre-prostatectomy population rather than on those going on to radiation therapy. This is, in large part, because the tissue obtained intraoperatively so readily avails itself to evaluation of immediate pathologic outcomes. Accordingly, early feasibility studies examined such endpoints as rate of positive surgical margins, change in tumor size by examination and imaging, pathologic stage, and biochemical response to therapy. Most studies favorably reported clinical responses to preoperative hormonal therapy, both in regard to PSA response and pre- and postoperative measures of tumor volume (pathologic downstaging).^18,19

Based on these results, multiple phase III trials have been performed comparing neoadjuvant androgen deprivation plus prostatectomy with surgery alone.^20–27 Although each trial varied slightly as to the definition of high risk, most patients were stages T2–T3. The duration of androgen deprivation, as in the above feasibility trials, was usually 3 months, and reports concentrated, at least initially, on immediate postoperative outcomes. These early analyses confirmed the findings of the phase II studies, with significantly increased rates of organ-confined disease, decrease in positive surgical margins, and decrease in pathologic stage compared with the control cases.^20,21,22,27 Subset analysis in several of the studies found that these improvements were consistently significant in the T2 population, with less convincing data for patients with T3 disease.^22,23 Despite the optimism that this downstaging might translate into clinically significant benefit, however, analyses of the long-term endpoints of these studies were disappointing (Table 2).^23–26 With follow-up ranging up to 7 years, no difference was seen in either biochemical relapse-free, or overall, survival. Only one trial, that by Schulman et al,²³ found local recurrence to be decreased, and this was only in the subset of patients with T2 disease.

Table 2.

Neoadjuvant Hormonal Therapy Plus Prostatectomy: Long-Term Results

	Patients		Maximum
Trial	(No. and Stage)	Treatment	Follow-Up	Outcome
Klotz et al.24	n = 213	3 mo CPA, then RP	3y	Lower rate positive margins
	T1b-T2c	vs RP alone		(27.7% vs 64.8%, P= .001)
				No difference in PSA relapse
				(40.2% vs 30.1%, P=0.3233)
Soloway et al.25	n = 282	3 mo L + F, then RP	5y	Lower rate positive margins
	T2b	vs RP alone		(18% vs 48%, P < .001)
				No difference in PSA relapse
				(35.2% vs 32.4%, P = .663)
Schulman et al.23	n = 402	3 mo G + F, then RP	4y	Pathologic downstaging T2 only
	T2–T3	vs RP alone		(8% vs 3%, P < .01)
				Improved local recurrence T2 only
				(3% vs 11%, P = .03)
				No difference in PSA relapse
				(26% vs 33%, P = .18)
Aus et al.26	n = 126	3 mo T, then RP	7y	Lower positive surgical margins
		vs RP alone		(23.6% vs 45.5%, P = .016)
				No difference PSA relapse
				(50.2% vs 48.5%, P = .588)

CPA, cyproterone acetate; RP, radical prostatectomy; L,leuprolide; G, goserelin; F, flutamide; T, triptorelin.

The discordance between the improvement in immediate pathologic endpoints and the lack of benefit in any clinically significant outcome is clear. This may imply that local prostate cancer cells may be more sensitive to the effect of androgen deprivation than are cells that have escaped the prostate, or that brief exposure to hormonal therapy is ineffective in eliminating a significant fraction of these cells. The findings in a study by Wood and colleagues provide additional support for this hypothesis.²⁷ In this case-control protocol, 60 patients with clinically localized prostate cancer considered to be at high risk for extraprostatic disease (cT2b-c, or stage T1c-2a with serum PSA ≥ 10 ng/mL) were analyzed for the presence of circulating prostate cancer cells with reverse transcriptase polymerase chain reaction amplification of the PSA mRNA in bone marrow samples. Thirty-one of these patients had received neoadjuvant hormonal therapy before radical prostatectomy, and 29 had prostatectomy alone. As with the above trials, statistically significant improvements were made in the rate of organ-confined disease with neoadjuvant therapy compared with prostatectomy alone. Although the absolute number of patients with organ-confined disease increased, it appeared that many of these still had disease outside the prostate. Specifically, 45% of neoadjuvant patients with organ-confined disease were found to have circulating cells, as opposed to only 6% of those with organ-confined disease who underwent surgery alone. This finding suggests that circulating cells (ie, those that go on to cause distant disease) may be biologically different and less sensitive to hormonal manipulation than are cancer cells within the prostate.

Neoadjuvant Hormonal Therapy and Radiotherapy

Compared with the disappointing results with neoadjuvant hormonal therapy in pre-prostatectomy patients, studies involving neoadjuvant androgen deprivation with radiotherapy seem more promising. Three phase III trials have been conducted evaluating androgen blockade of various types before radiation therapy (Table 3).^28–31 Although consistent improvement was seen in disease-free survival, only the Radiation Therapy Oncology Group (RTOG) study 86-10 was powered to detect a difference in overall survival.²⁹ No significant improvement was achieved in overall survival in all patients receiving combined modality therapy (50.5% vs 41%, P = .10), but there was a significantly better overall survival in the subset of patients with Gleason scores 2–6. Salvage hormonal therapy for treatment patients who eventually relapsed was not compromised.²⁹

Table 3.

Phase III Trials with Neoadjuvant Hormonal Therapy and Radiotherapy

Trial	Patients (N)	Therapy	Follow-up	Outcome
RTOG86–1028,29	456	G + F 2 mo prior	Median 8.6 y	Local control: 42% vs 30%, P = .016
		to and during RT		↓ distant mets: 35% vs 45%, P = .04
		vs RT alone		↑ DFS: 33% vs 21%, P =.004
				↓ cause-specific mortality:
				23% vs 31%, P = .05
				↑ overall survival for Gleason 2–6: 66.2% vs 41.4%, P =.015
				66.2% vs 41.4%, P = .015
Porter et al.30	208	12 wk CPA	Up to 3 y	↑ clinical DFS:
				70.6% vs 48.7%, P = .019
				↑ biochemical DFS:
				47.4% vs 21.5%, P = .001
Laverdiere et al.31	120	3 mo CAB, then RT		↓ residual cancer at bx at 12, 24 mo
		vs CAB 3 mo prior to,
		during, 6 mo after RT
		vs RT alone

G, goserelin; F, flutamide; CPA, cyproterone acetate; RT, radiotherapy; DFS, disease-free survival; CAB, combined androgen blockade; mets, metastases; bx, biopsy.

It appears, therefore, that there may be real clinical benefit in time to progression with hormonal therapy before definitive radiation. The improvement in overall survival appreciated in patients with a Gleason score of 2–6 is interesting and a bit counterintuitive, but this finding needs to be verified with further studies. Why a benefit is seen in patients undergoing radiation therapy and not prostatectomy is unclear. Some insight into this may be provided by data from animal studies performed by Zietman and colleagues, ³² showing that neoadjuvant androgen deprivation may enhance the effects of radiation. Overall, however, the data is encouraging, and efforts continue to be made to identify ways in which to maximize the benefits seen.

Impact of Duration of Therapy

One factor that has affected the outcome of the above trials is the duration of androgen deprivation. Two randomized trials treated men with either the standard 3 months of neoadjuvant combined androgen blockade or prolonged combined androgen blockade (6 months or 8 months, respectively).^33,34 Both studies found that increased length of hormonal therapy resulted in further reduction in tumor volume. The RTOG trial 92-02 took this concept one step farther and randomized 1554 men treated with flutamide (Eulexin®, Schering Plough, Kenilworth, NJ) and goserelin (Zoladex®, AstraZeneca, Wilmington, DE) for 2 months before and during radiation to an additional 24 months of androgen deprivation with goserelin or observation.³⁵ Improvements were noted in disease-free survival, local progression, and biochemical failure. Although in the whole group analysis no overall survival benefit was seen, the subset of patients with Gleason scores of 8–10 had a significantly improved overall survival (80% vs 69%, P = .02). It is possible, therefore, that with longer duration of androgen deprivation, neoadjuvant hormonal therapy might have shown stronger results in the above trials.

Neoadjuvant Chemotherapy

Historically it was felt that, at diagnosis, prostate cancer cells were inherently hormone-dependent and that the development of androgenindependence was a progression over time, at least in part owing to the pressure of hormonal therapy. Evidence is increasing, however, that prostate cancer cells are, from the start, a heterogeneous population, with varying degrees of hormone sensitivity. This provides the theoretical foundation for the utility of cytotoxic therapy in the early-stage setting. As noted above, although the historic paucity of active chemotherapeutic agents in prostate cancer has dampened enthusiasm for this approach, the utility of systemic chemotherapy of prostate cancer is dramatically increasing.

Chemotherapy in Advanced Prostate Cancer

Before the 1990s, overall response rates with available chemotherapeutic agents averaged less than 10%.³⁶ Mitoxantrone, (Novantrone®, Serono, Inc., Geneva, Switzerland), an anthracycline derivative, was one of the first agents to have an established role in treatment of advanced disease. Although biochemical response rates to mitoxantrone therapy, when combined with steroids in symptomatic hormone-refractory patients, have been just over 20% without improvement in overall survival, clinically significant benefit has been achieved in both symptoms and progression-free survival when compared with steroids alone.³⁷

Another, older agent with some activity in prostate cancer is estramustine (Emcyt®, Pharmacia and Upjohn, Kalamazoo, MI), a synthetic compound with both estrogenic and nitrogen mustard properties. As a single agent, response rates are suboptimal. In phase II trials combined with such agents as etoposide (VePesid®, Bristol-Myers Squibb Company, New York), vinca alkaloids, or the taxanes, however, biochemical and radiographic response rates have ranged from 45% to 67% and 32% to 45%, respectively.³⁶ This additive antitumor activity is felt to be due, in part, to the fact that estramustine targets the microtubule-associated proteins α- and β-tubulin at a site near but not overlapping the taxane site, theoretically providing complementary antitumor activity. In addition, these combination therapies have had attractive side-effect profiles and are generally very well tolerated. Several phase III trials of estramustine-based combinations have been conducted, with promising results. Hudes and colleagues³⁸ compared vinblastine (Velbe®, Eli Lilly Australia, West Ryde, New South Wales) plus estramustine with vinblastine alone and found statistically significant improvements in both progression-free survival and overall survival, with a median survival of 12.5 months for the combination, compared with 9.4 months for vinblastine alone (P = .051). A phase III trial by the Southwest Oncology Group comparing docetaxel (Taxotere®, Aventis Pharmaceuticals, Bridgewater, NJ) and estramustine to mitoxantrone plus prednisone has completed accrual, and results are pending.³⁹ The results of this trial should further clarify the role of such combination therapy in patients with advanced disease.

Trials Investigating Neoadjuvant Chemotherapy

Based on the activity of agents in advanced disease and the above-noted advantage to neoadjuvant therapy in easily establishing feasibility and efficacy, chemotherapy is being studied in the neoadjuvant setting. Although there are no phase III trials investigating this approach, early trials provide promising data (Table 4).^40–44 Despite minor differences in the endpoints measured and in the agents used, several similar conclusions emerge. Specifically, these therapies clearly result in decreased disease burden—clinically, pathologically, and biochemically. Furthermore, these agents were very well tolerated, and the disease-free survival rates were promising. Whether these results will translate into clinically significant benefit in a phase III setting, however, remains to be seen.

Table 4.

Neoadjuvant Chemotherapy

Trial	Regimen	Local Therapy	N	Outcome
Clark et al.40	Estramustine + etoposide	Surgery	18	DFS at 1 y 88%
				No pCR
				Organ-confined 31%
Pettaway et al.41	KAVE + CAB	Surgery	27	DFS at 1 y 69%
				No pCR
				Organ-confined 33%
Ben-Josef et al.42	Estramustine + etoposide	Radiotherapy + estramustine	18	DFS at 3 y 73%
Khil et al.43	Estramustine + vinblastine	Radiotherapy	65	DFS at 5 y 48%
Zelefsky et al.44	Estramustine + vinblastine	Radiotherapy	27	DFS at 2 y 60%
Hussain et al.4	Estramustine + docetaxel	Either surgery or radiotherapy	21	No pCR
				Negative margins 70%
				No node-positive disease

DFS, Disease-free survival, including biochemical recurrence; pCR, pathologic complete response; CAB, combined androgen blockade; KAVE, ketoconazole, doxorubicin, vinblastine, estramustine.

As noted above, duration of hormonal effect, including that of estramustine, may have a significant impact on outcome. In an effort to maintain the efficacy of combination therapy but minimize the hormonal effect, Hussain and associates studied the combination of docetaxel and shorter (3-day) duration of estramustine before either radiotherapy (n = 11) or prostatectomy (n = 10). Clinical response, as defined by a combination of DRE, imaging, and PSA level, was noted in all patients, and 8 of 11 patients who went on to radiation were assessed with post-chemotherapy biopsies. Although none of the post-prostatectomy patients achieved a pathologic complete response, pathologic changes consistent with antitumor effect were evident, and 2 of the post-radiation therapy patients had negative sextant biopsies. Interestingly enough, despite the shorter estramustine duration, all patients had suppressed testosterone levels.

Newer approaches are building on the increasing knowledge regarding these agents and how they interact to enhance antitumor activity. For instance, a recently activated trial at the University of Michigan is investigating the combination of docetaxel and capecitabine (Xeloda™, Roche Pharmaceuticals, Nutley, NJ) in the neoadjuvant setting. The rationale for this combination is based on extensive preclinical data, which indicate that docetaxel induces a high level of thymidine phosphorylase, the enzyme responsible for metabolizing capecitabine into the active form 5-fluorouracil in tissues, resulting in synergistic effects in nude mice.⁴⁵ This enzyme has been shown to be a tumor-associated angiogenesis factor with much higher concentration in tumor cells than in normal cells, providing some tumor selectivity.⁴⁶

Hopefully, by capitalizing on novel preclinical data and our experience with these agents in other tumor types, we will find combinations that provide prostate cancer patients with new options to maximize their odds of cure. At this point, however, neoadjuvant chemotherapy or chemohormonal therapy should be considered experimental in prostate cancer.

Conclusion

The genitourinary oncology community is clearly beginning to appreciate the need to incorporate chemotherapy early in the course of treatment for high-risk prostate cancer. The data described above suggest that neoadjuvant chemotherapy is feasible, is not associated with unexpected toxicities, and does not negatively impact the benefits or toxicities of local therapy. Perhaps more importantly, however, our experience thus far has raised some key questions:

1. What are the appropriate clinical, pathologic, and molecular endpoints in the phase II settings?

2. What are the appropriate imaging modalities and pathologic criteria to assess response?

3. Why do we not see P0 in prostate resections?

4. How do we distinguish a chemotherapy from a hormonal effect?

5. How do we optimally sequence the two systemic therapies to maximize efficacy?

6. What is the efficacy threshold that should trigger a phase III trial?

Using these questions both as a springboard for discussion within our multidisciplinary community and to initiate carefully designed clinical trials, we have the opportunity at hand to make broad strides in the treatment of prostate cancer.

Main Points.

• Despite adequate local therapy, 15%–40% of those patients diagnosed in the early stages of prostate cancer go on to develop progressive, usually systemic disease; these recurrences are likely due to micrometastatic foci present at the time of local therapy rather than to inadequate local techniques.

• Neoadjuvant therapy has been studied as a means of optimizing therapy to increase the number of men who are cured, theoretically through downstaging the disease and eradicating subclinical deposits at the earliest possible time.

• Multiple phase III trials comparing neoadjuvant androgen deprivation plus prostatectomy with prostatectomy alone found significantly increased rates of organ-confined disease and decreases in positive surgical margins, compared with the control cases; however, analyses of the long-term endpoints of these studies were disappointing.

• Compared with the disappointing results with neoadjuvant hormonal therapy in pre-prostatectomy patients, studies involving neoadjuvant androgen deprivation with radiotherapy seem more promising. Why a benefit is seen in patients undergoing radiation therapy and not prostatectomy is unclear.

• Chemotherapy is being studied in the neoadjuvant setting, and although there are no phase III trials investigating this approach, early trials provide promising data. Agents are very well tolerated.