Phase II trial of neoadjuvant docetaxel and gefitinib followed by radical prostatectomy in patients with high-risk, locally advanced prostate cancer




Prostate cancer trials investigating neoadjuvant hormonal therapy, followed by surgery, have demonstrated that elimination of all tumor cells from the primary site is rare. The authors report a phase 2 trial assessing the efficacy and toxicity of docetaxel and gefitinib in patients with high-risk localized prostate cancer as neoadjuvant therapy before radical prostatectomy (RP).


Thirty-one patients with high-risk prostate cancer were treated with docetaxel and gefitinib for 2 months before RP. All patients met the criteria of clinical stage T2b-3 or serum prostate-specific antigen (PSA) level >20 ng/mL, or Gleason score of 8 to 10. The primary endpoint was pathologic complete response. Secondary objectives included clinical response. When available, endorectal coil magnetic resonance imaging (eMRI) was performed as part of clinical response evaluation. Immunohistochemical staining of epidermal growth factor receptor and HER-2/neu was performed on prechemotherapy and postchemotherapy prostate tissue.


The median age of the patients was 60 years, the median pretreatment PSA level was 7.43 ng/mL, and the median Gleason score was 8. Clinical staging prior to treatment consisted of: T1 in 4 patients, T2 in 17 patients, and T3 in 10 patients. One patient with enlarged pelvic adenopathy and T4 disease did not undergo RP. Thirty patients received all scheduled therapies including RP. Grade 3 toxicities included asymptomatic liver function test elevation in 4 (13%) patients, diarrhea in 1 (3%) patient, and fatigue in 1 (3%) patient. One patient experienced grade 4 toxicity with elevated alanine aminotransferase. RP specimen pathology demonstrated residual carcinoma in all cases. Twenty-nine (94%) patients achieved a clinical partial response, including 35% of patients who demonstrated radiographic improvement on eMRI.


No pathologic complete response was noted in 31 patients treated with docetaxel and gefitinib. This combination was well tolerated, and did not result in increased surgical morbidity. Cancer 2009. © 2009 American Cancer Society.

Biochemical or clinical failure for patients with high-risk locally advanced prostate cancer after radical prostatectomy (RP) is estimated to be 50% to 90% within 5 to 10 years after definitive surgical therapy alone.1-3 Effective therapy for high-risk locally advanced prostate cancer is needed. Trials investigating neoadjuvant hormonal or chemotherapy, followed by surgery, demonstrate that the elimination of all tumor cells in the primary site is rare.4-10 Prostate cancers demonstrate heterogeneity, even in early stages of disease, for a variety of biomarkers, including platelet-derived growth factor, c-kit, and epidermal growth factor receptor (EGFR).11 The addition of novel targeted therapies, in combination with chemotherapy or alone, is a potentially promising approach to neoadjuvant therapy for high-risk prostate cancer.

Members of the EGFR family have been identified as potential drug targets that are differentially expressed on tumor cells after androgen ablation.11 Gefitinib, a small-molecule tyrosine kinase inhibitor of EGFR, has demonstrated antiproliferative activity against prostate cancer cell lines.12 Docetaxel has shown efficacy and improvement in survival in patients with metastatic prostate cancer, and the combination of taxane and gefitinib has produced regression in prostate cancer cell lines.13, 14 We performed a phase 2 trial to assess the pathologic complete response (pCR) in 29 evaluable patients using the combination of docetaxel and gefitinib as neoadjuvant therapy before RP in patients with high-risk prostate cancer.



Between July 2003 and January 2006, 31 patients with high-risk prostate cancer entered this phase 2 trial approved by the institutional review board at Virginia Mason Medical Center. Eligibility criteria included histologically confirmed locally advanced adenocarcinoma of the prostate; clinical stage T2b-3 disease or a serum prostate-specific antigen (PSA) ≥20 ng/mL, or a Gleason sum score of 8 to 10. In addition, clinical T2 patients were eligible if endorectal coil magnetic resonance imaging (eMRI) demonstrated T3 disease or a Gleason sum score of 4 + 3 cancer in ≥5 biopsies (with a minimum of 10 biopsies), informed consent, and no evidence of metastatic disease confirmed with computed tomography, and bone scan performed within 60 days of enrollment. An Eastern Cooperative Oncology Group (ECOG) performance status of 0 to 2 and adequate hematologic, renal, and hepatic function (which included an absolute granulocyte count ≥1500 cells/mm3, a platelet count ≥100,000 cells/mm3, serum bilirubin within 1.5× the upper limit of normal [ULN], transaminase levels ≤2.5× ULN, and a serum creatinine ≤1.5× ULN) were required for enrollment. Exclusion criteria included existing peripheral neuropathy ≥grade 1; concurrent or prior treatment with hormonal therapy (excluding finasteride for obstructive voiding symptoms), radiation, or biologic therapy for prostate cancer; prior malignancy; and the concomitant use of phenytoin, carbamazepine, barbiturates, rifampicin, or St. John's wort. Patients unwilling to use effective means of contraception were excluded; contraception was continued for 3 months after treatment.

Treatment Schedule

The dose and schedule of docetaxel and gefitinib in the current study was derived from lung cancer trials.15, 16 Gefitinib was supplied by AstraZeneca Pharmaceuticals (London, England) as 250-mg–strength, film-coated tablets. The dose of 250 mg/day was administered orally beginning the first day of docetaxel chemotherapy. Docetaxel was administered weekly at a dose of 36 mg/m2 over 1 hour with dexamethasone before and after chemotherapy. Premedication with dexamethasone was given to reduce the incidence and severity of fluid retention as well as the severity of hypersensitivity reactions. Dexamethasone at a dose of 4 mg orally was given the night before, the morning of, and the evening after docetaxel administration. Docetaxel was given weekly for 3 weeks with 1 week off for 2 cycles, with 1 cycle constituting 28 days. Gefitinib was administered orally continuously for 56 days. Neoadjuvant hormonal therapy was not provided, given previous data demonstrating its limited role in this setting, and to clarify the influence of docetaxel and gefitinib without confounding variables.

Dose Modifications and Supportive Care

Because docetaxel and gefitinib have different toxicities, we judged which was the causative agent and reduced accordingly. Toxicity was graded according to the National Cancer Institute Common Toxicity Criteria v. 3.0 grading system. In the case of hematologic toxicity, grade 4 thrombocytopenia required a docetaxel dose reduction of 25%; if a patient experienced a grade 3 or 4 neutropenia, treatment was withheld until the toxicity was resolved to grade 1 or less. No dose reductions were required for anemia. The routine use of granulocyte–colony–stimulating factor (GCSF) was not allowed. In patients who experienced neutropenic fever, the use of GCSF was allowed, and used at the discretion of the investigator. No patients required the use of GCSF or erythropoietin-α.

Liver function tests (LFTs) were evaluated every week while patients were receiving chemotherapy. Patients who developed abnormal liver function tests for any reason while on study, had the following modifications: Docetaxel was reduced by 25% for total bilirubin >2 mg/dL and aspartate aminotransferase (AST) >ULN. Docetaxel was held in patients whose bilirubin was ≥2.5 mg/dL or who had alkaline phosphatase or AST levels >5× ULN. Docetaxel therapy was resumed at a 25% dose reduction in patients whose LFTs recovered within 3 weeks to grade 1 or less. Patients who received dose reduction for docetaxel were allowed to be re-escalated once LFTs returned to normal. Patients whose LFTs did not recover within 3 weeks of dose delay or reduction were taken off study.

In the case of nonhematologic toxicity from gefitinib, only grade 3 to 4 diarrhea was considered significant to withhold gefitinib therapy and reinstituted when toxicity returned to grade 2 or less. Patients with poor tolerance to skin toxicity were allowed a 14-day interruption and were then allowed to resume at the 250-mg dosing after resolution of rash. Many patients taking part in other similar trials experienced less severe rashes when treatment was resumed.

Patients were monitored weekly with physical examinations, complete blood counts, and serum chemistry. In addition, a digital rectal examination, serum PSA, and testosterone level were performed at screening, at Week 5, and at the completion of docetaxel and gefitinib. Each patient was reassessed by their primary surgeon for RP after 2 cycles of chemotherapy. RP was performed within 3 to 5 weeks of the last week of chemotherapy.

Statistical Analysis

Several studies of neoadjuvant chemotherapy before RP have been conducted in patients with high-risk, clinically localized prostate cancer, and all have reported pathologic response rates of 0%.6-10 Thus, a pathologic response of ≥10% in the current study would be a significant finding. With 29 patients, there was a 95% probability of observing at least 1 responder, if the underlying response rate is 10%. Similarly, if no successes were observed in 29 patients, a 90% confidence interval (CI) would not contain 10%. Upon completion of the study, the true pCR rate was estimated based on RP specimens, and an exact CI was constructed for the 29 patients. The target accrual was 29 evaluable patients, a sample size associated with a 95% CI for the pathologic response rate of ±10%.

Radical Prostatectomy

All RPs were performed in an open fashion through an 11-cm lower midline incision as described by Walsh and Donker.17 All patients underwent a staging pelvic lymph node dissection. Margins of resection included the obturator nerve, external iliac vein, Cooper ligament, and bifurcation of the external and internal iliac arteries. In the absence of macroscopically metastatic disease, the prostate was exposed by entering the endopelvic fascia. The deep dorsal venous complex was divided and ligated to expose the prostatic apex and the prostatic fascia. Nerve-sparing techniques, via incision of the lateral prostatic fascia, were used only in the absence of clinical evidence of extracapsular disease extension. Lateral pedicles were divided, and the seminal vesicles were dissected completely. Finally, the prostate was removed in a nonbladder neck–sparing fashion. Anastomosis was performed using 6 monofilament 2-0 sutures.

Correlative Studies

EGFR and HER-2/neu were evaluated by immunohistochemical staining on pretherapy (biopsy) and post-therapy (RP) specimens to evaluate the relative changes in expression. Correlative studies were performed at PhenoPath laboratories for EGFR (Dako AO485, Glostrup, Denmark) and HER-2/neu (Zymed #28-005, South San Francisco, Calif) on unstained slides of both biopsy and prostatectomy specimens. A semiquantitative scoring system was used to report results. All pretreatment biopsies and RP specimens were reviewed by our reference pathologist (C.I.).


Patient Characteristics

Thirty-one patients were enrolled into this study (Table 1). The median age was 60 years (range, 46-74 years). The median ECOG performance status was 0. The median PSA was 7.43 ng/mL (range, 0.73 ng/mL-113 ng/mL), and the median Gleason score was 8 (range, 6-10). Clinical staging of the patients consisted of T1c in 4 patients, T2a in 9 patients, T2b in 4 patients, T2c in 4 patients, and T3 in 10 patients. On the basis of the Kattan nomogram,18 the predicted biochemical progression-free probability after prostatectomy was 50%.

Table 1. Patient Characteristics
  1. ECOG indicates Eastern Cooperative Oncology Group; PSA, prostate-specific antigen.

Median age, y60
ECOG performance status 
 030 (97%)
 11 (3%)
Clinical stage (2002 criteria) 
Biopsy Gleason score (pretreatment) 

Clinical Response

One patient completed 2 cycles of chemotherapy, but was found to have an enlarged pelvic lymph node and T4 disease at the time of surgical exploration, and did not undergo RP. Postchemotherapy PSA reductions were observed in 21 (68%) patients with a range of reduction from 3% to 91%. PSA increases of >25% over baseline were noted in 2 patients. Testosterone levels monitored at baseline, mid-therapy, and postchemotherapy were stable throughout treatment with docetaxel and gefitinib. We were able to evaluate 26 patients with pre- and post-eMRI, and in 9 (35%) patients, radiographic changes in tissue were demonstrated.


Thirty-one patients were evaluable for toxicity from chemotherapy. Over the course of treatment, the side effect profile was mild in intensity. Toxicities included fatigue, diarrhea, abdominal cramping, mouth sore, anemia, and asymptomatic elevation in liver function enzymes (Table 2). Grade 2 and 3 fatigue occurred in 6 and 1 (22%) patients. Grade 2, 3, and 4 asymptomatic elevations in liver function enzymes were observed in 5, 4, and 1 (32%) patients, respectively. Three (10%) patients experienced grade 2 or 3 diarrhea. Twenty-nine patients had no delay in time to RP. In 2 patients, RP was delayed 8 to 10 weeks after chemotherapy because of pneumonia (no hospitalization), and 1 patient experienced hematuria during the last cycle of chemotherapy. This patient underwent cystoscopy, with examination performed under anesthesia, with findings of diffuse erythema in the bladder. Biopsies revealed severe bladder mucosal inflammation with cystitis glandularis, which was believed to possibly be related to chemotherapy. RP was performed without complication, and the patient had had no recurrence of hematuria at the time of last follow–up.

Table 2. Toxicity by Common Criteria*
ToxicityNo. of Patients With Grade 2No. of Patients With Grade 3No. of Patients With Grade 4
  • LFTs indicates liver function tests.

  • *

    Grading determined according to the National Cancer Institute Common Toxicity Criteria v. 3.0 grading system.

  • Worsening from pre-existing condition.

Constitutional symptoms
 Abdominal cramping200
 Mouth sore/tongue200
 Elevated LFTs541
Bone marrow   

Pathologic Outcome

The 30 patients who underwent prostatectomy had a median Gleason sum score of 7 (range, 6-9) (Table 3). One patient underwent surgical exploration and did not complete RP and was not included in the pathologic outcome. Postprostatectomy pathologic stage consisted of pT2 in 17 patients, pT3 in 13 patients, and N1 disease in 4 patients (Table 3). Pathologic grade included a Gleason score of 6 in 2 patients, a Gleason score of 7 in 19 patients, a Gleason score of 8 in 1 patient, and a Gleason score of 9 in 8 patients. The median volume of residual disease consisted of 16.5% (range, <1%-90%) cancer in the specimen, with a mean of 26%. This translated into a range of <1 g to 49 g of residual carcinoma. Thirteen (43%) patients had extracapsular extension, and 8 (27%) had seminal vesicle involvement. Staging pelvic lymph node dissection yielded a median of 5 lymph nodes, with a range of 2 to 13 lymph nodes. Four (13%) patients had microscopic lymph node involvement. The median PSA after prostatectomy was 0.01 ng/mL. Undetectable PSA (defined as <0.01 ng/mL) was achieved in 16 (53%) patients for the first PSA drawn 6 weeks to 3 months after RP. Twenty-five (83%) patients achieved PSA of <0.2 ng/mL.

Table 3. Pathology Outcome (n = 30)
  1. PSA indicates prostate-specific antigen.

Pathologic stage 
Lymph node positive4 (13%)
Surgical Gleason score 
Extracapsular extension13 (43%)
Seminal vesicle invasion8 (27%)
Negative surgical margin20 (67%)
PSA <0.2 ng/mL25 (83%)


As an exploratory objective, we performed immunohistochemical staining on 29 patients with HER-2/neu and EGFR using a reference laboratory (PhenoPath) on biopsy specimens and prostates postprostatectomy. By using immunohistochemical staining criteria developed for breast cancer, the majority of patients stained negative for HER-2/neu, and this did not change after therapy. In EGFR staining, the majority (62% before and 70% after therapy) of patients stained positive for EGFR, but there was no appreciable changes in pre- and postprostatectomy specimens.

Surgical Outcomes

Surgical outcomes were divided into intraoperative, postoperative, and long-term outcomes (Table 4). The estimated blood loss in the current series was similar to other prospective series of patients undergoing neoadjuvant chemotherapy, with a median blood loss of 575 mL (mean, 768 mL; range, 100 mL-2000 mL).6, 9 A single rectal injury occurred. There were no thromboembolic events. Late events included 2 inguinal hernia repairs within 1 year of RP. Potency rates, as defined by International Index of Erectile Function scores, and continence rates are reported in Table 4. These quality of life outcomes were similar to our series of unilateral nerve-sparing RP performed in patients not undergoing chemotherapy before surgery.

Table 4. Surgical Outcomes
  • AUA indicates American Urological Association; IIEF, International Index of Erectile Function; RPR, radical prostatectomy.

  • *

    In patients undergoing unilateral nerve-sparing procedure.

 Estimated blood loss, cc575 (range, 100-2000)30
 Intraoperative complications  
  Rectal injury1 (3%)30
 Median length of hospital stay, d2 (range, 1-4)30
 Symptomatic lymphocele030
 Bladder neck contracture030
Long term  
 Continence25/30 (83%)30
 Mean AUA quality-of-life score3 (range, 0-6)17
 Potency (IIEF)*7.5 (range, 2-17)15
 Postoperative morbidity  
  Inguinal hernia repair within 1 y RPR2 (7%)30


At the time of last follow–up, all patients had completed protocol therapy, and the median duration of follow-up was 28 months (range, 10 months-42 months) from RP. The 1 patient who was found to be inoperable progressed to metastatic disease after combined modality therapy with radiotherapy and hormonal therapy. This patient died 24 months after study enrollment. Eight (27%) patients underwent additional hormonal and radiotherapy, or hormonal therapy alone, because of lymph node–positive disease discovered at the time of surgery or a postoperative PSA level of >0.4 ng/mL, at the discretion of the treating physician. Of the remaining 22 patients, 20 (66%) were determined to be biochemically free of disease (PSA level <0.01 ng/mL) with no additional therapy.


Prompted by reports that neoadjuvant chemotherapy with docetaxel for prostate cancer was safe and feasible,6-10 that treatment with docetaxel in castrate progressive metastatic prostate cancer improved survival,19, 20 and that the combination of docetaxel and gefitinib had synergistic properties, this phase 2 trial of gefitinib and docetaxel was undertaken in patients with high-risk localized prostate cancer before RP. We observed no pCR in the 31 patients treated. This combination was well tolerated, and did not result in increased surgical morbidity.

Since the first study of neoadjuvant therapy in locally advanced prostate cancer, our trial and 3 others have investigated either docetaxel alone or in combination with other chemotherapy agents in a comparable patient population (Table 5).6, 8, 9, 21 Several issues are clear from these series of trials. Although all used docetaxel alone or combined with other agents, collectively there was no pCR. In addition, the optimal regimen and the duration of neoadjuvant therapy necessary to produce a pathologic response remains unclear. Surgical morbidity was acceptable across all studies.

Table 5. Summary of Published Trials of Neoadjuvant Docetaxel Prior to Radical Prostatectomy
 Current SeriesFebbo 20058Dreicer 20049Beer 20046
  • pCR indicates pathologic complete response; NA, not applicable; ECE, extracapsular extension; PSA, prostate-specific antigen.

  • *

    Organ confined, indicates T2a-cNO.

No. of patients30162821
RegimenDocetaxel, gefitinibDocetaxelDocetaxelDocetaxel, mitoxantrone
Localized disease*57%38%11%43%
Lymph node positive13%0%14%29%
Negative surgical margin67%NA70%76%
Seminal vesicle invasion27%50%32%27%
PSA <0.2 ng/mL83%NA93%67%

We tested the hypothesis that a noncytotoxic, tyrosine kinase inhibitor of EGFR combined with docetaxel would achieve a pCR in patients with prostate cancer. Abnormally elevated EGFR tyrosine kinase activity is observed in 40% to 90% of prostate cancers.22 However, to our knowledge no activating mutations have been described in prostate cancer.22 Gefitinib, by inhibiting proliferation, inducing apoptosis, and promoting drug delivery, could potentially enhance the cytotoxic effect and act synergistically with docetaxel. By administering therapy in the neoadjuvant setting, we were able to determine whether there were any effects of therapy on prostate tissue. We did not find any correlation in the 2 markers we tested, EGFR and HER-2/neu. There may be several reasons for this, including that the exposure of gefitinib may be too brief to see changes detectable by immunohistochemistry, and that the methods used may not be sensitive enough to detect a difference before and after prostatectomy.

Since this trial was designed and completed, reports of gefitinib alone in metastatic and nonmetastatic castratory hormone progressive prostate cancer have found no responses (as defined as PSA declines).23-26 In addition, there were no responses to the combination of docetaxel and gefitinib compared with the treatment of docetaxel alone.26 Furthermore, although there were no changes in the pharmacokinetics of docetaxel combined with 250 mg of gefitinib, at a higher dose, the area under the curve concentration for docetaxel was reduced by 20%.26 Most of the patients on this trial received a total dose of 500 mg of gefitinib before any dose reductions were performed.

Our trial evaluated 8 weeks of neoadjuvant therapy. Other trials have evaluated the same duration, and up to 6 months.8, 9, 21 To the best of our knowledge, the optimal duration and regimen for neoadjuvant chemotherapy of prostate cancer has not been established to date. Even with other trials investigating 6 months of neoadjuvant chemotherapy, there were no pCRs observed.21 As an example of how the regimen and duration of therapy impacts pCR, in trials of neoadjuvant for breast cancer, 4 cycles of chemotherapy reportedly achieves pCR ranging from 4% to 13%.27 With the addition of sequential therapy to a noncross-resistant regimen and up to 6 months of therapy, the pCR rate was reported to be 26% to 31%.27

The pathologic outcome and surgical morbidity in the current study were similar to other high-risk series.8, 21 In this study, 1 patient experienced rectal injury, which is similar to the rate of 4.2% reported from a similar series of high-risk patients without preoperative therapy.4 The positive surgical margin rate of 33% in the current series is similar to the rate of 48% to 65% reported in the literature for patients treated with hormonal therapy before RP.4, 5 Lymph node involvement, positive surgical margins, and organ-confined disease were comparable across the studies.

The concept of neoadjuvant therapy before RP is appealing, and bears further evaluation in randomized trials evaluating patients with high-risk, clinically localized prostate cancer comparing docetaxel and androgen ablation followed by RP with RP alone. Given the negative results reported in trials of neoadjuvant therapy and in the advanced setting, gefitinib does not appear to have activity either as a single agent or in combination with chemotherapy in prostate cancer.


We thank Nancy Tyler for administrative support and Dalia Kamarauskiene, Marisa Cody, and Rebecca Olson for additional research support.

Conflict of Interest Disclosures

Supported by Sanofi-Aventis and AstraZeneca Corporation.