Treatment of high-risk localized prostate cancer remains inadequate. The authors performed a phase 2 multicenter trial of neoadjuvant docetaxel plus bevacizumab before radical prostatectomy.
Treatment of high-risk localized prostate cancer remains inadequate. The authors performed a phase 2 multicenter trial of neoadjuvant docetaxel plus bevacizumab before radical prostatectomy.
Eligibility included any of the following: prostate-specific antigen (PSA) >20 ng/mL or PSA velocity >2 ng/mL/y, cT3 disease, any biopsy Gleason score 8 to 10, and Gleason score 7 with T3 disease by endorectal magnetic resonance imaging (MRI) at 1.5 T. Also, those with ≥50% biopsy cores involved and either Gleason score 7, PSA >10, or cT2 disease were eligible. Patients were treated with docetaxel 70 mg/m2 every 3 weeks for 6 cycles and bevacizumab 15 mg/m2 every 3 weeks for 5 cycles. The primary endpoint was partial response by endorectal MRI.
Forty-one patients were treated. Median age was 55 years (range, 40-66 years). Baseline characteristics included: median PSA, 10.1 ng/mL; cT2, 49%, cT3, 32%; and Gleason score 8 to 10, 73%. Thirty-eight of 41 (93%) patients completed all 6 cycles. Grade ≥3 adverse events were rare, although 3 of 41 (7%) experienced febrile neutropenia. Twelve patients (29%; 95% confidence interval [CI], 16%-45%) achieved a >50% reduction in tumor volume, and 9 patients (22%; 95% CI, 11%-38%) achieved a >50% post-treatment decline in PSA. Thirty-seven of the 41 patients underwent radical prostatectomy; there were no complete pathologic responses.
Neoadjuvant docetaxel and bevacizumab is safe, and results in reductions in both tumor volume and serum PSA, in men with high-risk localized prostate cancer. The role of neoadjuvant chemotherapy in prostate cancer, and perioperative antiangiogenic therapy in general, requires further elucidation through ongoing and planned trials. Cancer 2012. © 2012 American Cancer Society.
Prostate cancer (PCa) is the second leading cause of cancer death in men in the United States.1 Although radical prostatectomy (RP) is curative in the majority of patients with clinically localized PCa, 30% to 40% of men will develop prostate-specific antigen (PSA) recurrence, many of whom will ultimately die from their disease.2 As has been demonstrated in other solid tumors, the integration of perioperative systemic treatment may be necessary to increase the likelihood of cure in this high-risk subset of patients.
The successful development and implementation of perioperative systemic therapeutic strategies for the treatment of clinically localized solid tumors requires both an accurate means to risk stratify patients and the availability of active systemic therapy. Furthermore, intermediate endpoints that can be used in the phase 2 setting to help guide decisions regarding the initiation of large and lengthy definitive phase 3 trials are desirable. In patients with PCa, pretreatment clinical and pathologic variables, including Gleason score, PSA, and clinical stage, have been shown to accurately predict PSA recurrence and cancer-related death after RP.3, 4 Taxane-based chemotherapy demonstrates activity in a large proportion of patients with advanced disease as measured by post-treatment declines in PSA, symptomatic improvements, and prolonged survival.5-7 Finally, delivery of systemic therapy in the neoadjuvant setting affords the ability to assess response in the primary tumor, which has correlated with survival in many other solid tumors.
Several historical randomized trials exploring neoadjuvant androgen deprivation therapy before prostatectomy in patients with high-risk localized PCa failed to demonstrate an improvement in clinical outcomes. However, these trials were generally underpowered, suffered from short durations of treatment and follow-up, and did not use optimal risk stratification.8-12 The approval of docetaxel for the treatment of advanced PCa led to a renewed interest in the evaluation of neoadjuvant therapy in high-risk localized disease.13-15 We previously reported a phase 2 study of neoadjuvant docetaxel before RP in 19 men with high-risk localized PCa, demonstrating that treatment was well tolerated and associated with PSA declines of ≥50% in 58% of patients and tumor volume reductions of ≥50% by endorectal magnetic resonance imaging (MRI) in 21% of patients.15 On the basis of the results of our study and others, a phase 3 study of neoadjuvant chemotherapy plus androgen deprivation therapy followed by prostatectomy versus prostatectomy alone in patients with high-risk localized PCa was initiated by the Cancer and Leukemia Group B (CALGB 90203) and is ongoing.16
Bevacizumab is a humanized monoclonal antibody that binds to and neutralizes serum vascular endothelial growth factor (VEGF), a major mediator of tumor angiogenesis. The addition of bevacizumab to chemotherapy has led to improved clinical outcomes in patients with diverse solid tumors.17, 18 Phase 2 studies of docetaxel plus bevacizumab in castration-resistant PCa19-21 have demonstrated significant activity leading to a phase 3 CALGB trial in patients with metastatic disease.22 Recently, this study was reported not to demonstrate a survival advantage with the combination, but did show biological activity with a longer progression-free survival in the group receiving bevacizumab. Given that tumors generally cannot grow beyond 1 to 2 mm3 without recruitment of the neovasculature, the benefits of inhibiting VEGF may be more substantial in the setting of micrometastatic disease.23 Therefore, we initiated a phase 2 multicenter trial to evaluate the safety and efficacy of the combination of docetaxel and bevacizumab in men with high-risk localized PCa before RP.
This study was designed as a single-arm, 3-site (Dana Farber Cancer Institute, Beth Israel Deaconess Medical Center, Duke Comprehensive Cancer Center) phase 2 trial as part of the Prostate Cancer Clinical Trials Consortium.24 Eligible patients had histologic documentation of adenocarcinoma of the prostate and were candidates for RP. Patients were considered high risk for recurrence as defined by 1 or more of the following characteristics: Gleason score ≥8, Gleason score 7 and endorectal MRI T3 disease, clinical T stage of T3a or T3b, serum PSA ≥20 ng/mL, and PSA velocity of ≥2 ng/mL/y in the year before diagnosis. In addition, patients with ≥50% of the total number of biopsy cores positive for PCa were eligible provided that 1 of the following characteristics was also present: Gleason score of 7; clinical T stage of T2a, T2b, or T2c; and/or serum PSA ≥10 ng/mL. All patients were required to be free from evidence of metastatic disease, have an Eastern Cooperative Oncology Group performance status of 0 to 1, and have a serum testosterone of >100 ng/dL. Patients with a history of unstable angina, symptomatic peripheral vascular disease, New York Heart Association class ≥III heart failure, uncontrolled hypertension, myocardial infarction or stroke during the 12 months before enrollment, history of deep venous thrombosis or pulmonary embolism, known coagulopathy or bleeding diathesis, ongoing use of anticoagulant therapy, abdominal fistulas, gastrointestinal perforation, nonhealing ulcer or fracture, or spot urine protein:creatinine ratio >1.0 were excluded. Other eligibility criteria included adequate hematologic, hepatic, and renal function. The protocol was approved by the institutional review boards at each participating institution. Informed consent was obtained from all patients before enrollment.
The pretreatment evaluation included a complete medical history and physical examination including a digital rectal examination (DRE). Baseline studies included a spot urine protein:creatinine ratio, complete blood count with differential, serum chemistries including creatinine and liver function tests, PSA, testosterone, and an electrocardiogram. All patients underwent a baseline endorectal MRI at 1.5 T.
The treatment plan was as follows. All patients underwent a baseline endorectal MRI and a restaging endorectal MRI after 6 cycles of systemic therapy. Docetaxel was administered on day 1 at a dose of 70 mg/m2 intravenously every 21 days with standard dexamethasone premedications and antiemetics. Bevacizumab was administered on day 1 at a dose of 15 mg/kg intravenously every 21 days. Combination treatment with docetaxel plus bevacizumab was administered during cycles 1 to 5, and docetaxel alone was administered during cycle 6. Patients underwent a repeat DRE on cycle 2, day 1 and on cycle 4, day 1. Patients with disease progression by DRE underwent an early restaging endorectal MRI. Toxicity assessments were performed on the day of each treatment using the Common Toxicity Criteria of the National Cancer Institute version 3.0. The dose of docetaxel and/or bevacizumab was held and/or decreased as per an algorithm based on the specific toxicity.
Patients continued to receive therapy until completion of 6 cycles. Treatment was discontinued early in the event of progressive disease defined as at least 1 of the following: appearance of metastatic disease; serum PSA ≥150% of cycle 1 day 1 value, confirmed with another serum PSA taken at least 2 weeks later; and increase in tumor volume by at least 50% from baseline, as measured by endorectal MRI.
All patients were reevaluated by participating urologists for consideration of RP once chemotherapy was completed. The surgical approach, including decisions regarding lymphadenectomy and nerve-sparing techniques, was left to the discretion of the surgeon. The surgery was to take place between 5 and 8 weeks after completion of neoadjuvant therapy, allowing for full recovery from the last cycle of chemotherapy.
Antitumor response was assessed with an endorectal MRI repeated after completion of 6 cycles of neoadjuvant therapy. A response was defined as a decrease in tumor size of >50% for the largest lesion in the prostate by endorectal MRI using methodology as previously described.15 Briefly, on the baseline endorectal MRI, 1 target PCa lesion was identified. This lesion was identified on multiplanar T2-weighted imaging. The target lesion was identified as the largest of the lesions in the prostate and was ideally >0.5 mm3. The target lesion was measured and recorded by its longest diameter in 3 dimensions to derive a volumetric measurement. This measurement required a >50% reduction in size for the patient to be considered a responder. Serum PSA levels were checked at baseline, with each cycle of therapy, 4 to 6 weeks after RP, then at least every 6 months during the 5-year follow-up period. All RP specimens were submitted to a reference pathologist (M.B.) for central review.
This trial was designed as a Simon 2-stage single-arm phase 2 trial to differentiate an endorectal MRI response rate of ≥35% from a response rate of 18% (based on our prior study with docetaxel alone15) with a target alpha of ≤.06 and a beta of ≤.20. If 4 or more of the initial 17 patients responded, enrollment continued to a total of 42 patients. If 12 or more responses were observed, then the regimen was considered worthy of further study.
All patients who received at least 1 dose of protocol therapy were included in the assessment of response, regardless of their disease evaluations. Endorectal MRI and PSA response (the rate of any PSA decline and PSA decline by 50%) were reported with exact binomial 95% confidence interval (CI). Endorectal MRI and PSA response were also graphically described using waterfall diagrams.
The proportion of patients who achieved a nadir of <0.1 ng/mL post-RP was reported with 95% CI. Biochemical recurrence was defined as the first PSA ≥0.2 ng/mL (confirmed by subsequent PSA ≥0.2) or date of initiation of new therapy post-RP. Patients without biochemical recurrence were censored at last follow-up for PSA. Distribution and median time to biochemical recurrence were estimated using the Kaplan-Meier methodology.
Analyses were performed using SAS statistics software (SAS Institute Inc., Cary, NC).
An early stopping rule for toxicity was also used. After 17 patients underwent RP, their postoperative course was reviewed by a team of physicians including at least 1 urologist and 1 medical oncologist for toxicities that occurred that were both serious and unexpected. If 3 or more such unexpected grade 3 or 4 adverse events occurred, plans were to close the study. With this rule, the probability of detecting a true unexpected serious toxicity rate of 30% was 92.3%.
Between July 2006 and November 2008, 42 patients were enrolled on study; however, 1 patient withdrew from the study before receiving any treatment. Therefore, a total of 41 patients were included in the analysis. The disposition of these 41 patients is outlined in Figure 1. The baseline patient characteristics are presented in Table 1. The median age of enrolled patients was 55 years (range, 41-67 years). The majority of patients had cT2 (49%) or cT3 (32%) disease and a Gleason score of 8 to 10 (73%). The median baseline PSA was 10.1 ng/mL (range, 2.1-72.4 ng/mL).
|Median age, y (range)||55 (41-67)|
|Race, No. [%]|
|ECOG performance status, No. [%]|
|Clinical T stage, No. [%]|
|Clinical N stage, No. [%]a|
|Biopsy Gleason score, No. [%]|
|PSA velocity >2 ng/mL/y, No. [%]|
|Median baseline PSA, ng/mL (range)||10.1 (2.1-72.4)|
|Median baseline testosterone, ng/mL (range)||381 (187-962)|
Three of the 41 patients discontinued neoadjuvant therapy before completion of all 6 cycles (docetaxel-related allergic reaction = 1, patella fracture = 1, rising PSA = 1). The treatment delivery is outlined in Table 2. The median duration of protocol therapy (from day 1 cycle 1 until the date of RP) was 5.3 months (range, 1.4-6.3 months). Thirty-seven of the 41 patients underwent RP. The reasons for not proceeding with surgery in the remaining 4 patients included a Foley catheter-related puncture of the bladder neck and rectum before surgery, surgeon's decision based on results of lymphadenectomy, and withdrawal of consent in 2 patients. The median time from completion of chemotherapy to RP was 1.2 months (maximum, 1.8 months).
|Number of patients completing specified number of cycles|
|Completed 3 cycles||2|
|Completed 4 cycles||1|
|Completed 6 cycles||38|
|Number of docetaxel dose reductions||3|
|Number of docetaxel dose interruptions||8|
|Number of bevacizumab dose reductions||0|
|Number of bevacizumab dose interruptions||5|
The treatment-related adverse events experienced by >10% of patient are outlined in Table 3. The vast majority of adverse events were low grade and consistent with previously described side effects of docetaxel and bevacizumab. Notably, there were no episodes of grade 3 or 4 hypertension, bleeding, gastrointestinal perforation, or thromboembolic events. There were 3 episodes of grade 3 febrile neutropenia.
|Allergic reaction||3 (7%)||2 (5%)||1 (2%)||0|
|Anorexia||6 (15%)||2 (5%)||0||0|
|Anxiety||3 (7%)||1 (2%)||0||0|
|Blurred vision||7 (17%)||0||0||0|
|Cough||5 (12%)||0||1 (2%)||0|
|Diarrhea||7 (17%)||5 (12%)||0||0|
|Elevated AST||7 (17%)||1 (2%)||0||0|
|Epiphora||13 (32%)||2 (5%)||0||0|
|Epistaxis||20 (49%)||2 (5%)||0||0|
|Fatigue||21 (51%)||12 (29%)||0||0|
|Febrile neutropenia||0||0||0||3 (7%)|
|Fever w/o neutropenia||6 (15%)||0||0||0|
|Headache||8 (20%)||1 (2%)||0||0|
|Hiccoughs||2 (5%)||2 (5%)||0||0|
|Hyperglycemia||13 (32%)||24 (59%)||1 (2%)||0|
|Hypertension||3 (7%)||2 (5%)||0||0|
|Joint pain||6 (15%)||1 (2%)||0||0|
|Lymphopenia||4 (10%)||0||3 (7%)||0|
|Mucositis||10 (24%)||1 (2%)||0||0|
|Nail changes||9 (22%)||2 (5%)||0||0|
|Nausea||13 (32%)||2 (5%)||0||0|
|Neutropenia||1 (2%)||0||0||3 (7%)|
|Rhinitis||10 (24%)||2 (5%)||0||0|
|Voice changes||7 (17%)||0||0||0|
The median procedural time for the 37 patients undergoing RP was 240 minutes (range, 119-308 minutes), and median blood loss was 500 mL (range, 75-1700 mL). Two patients experienced grade 2 to 4 adverse events that extended hospitalization; both were grade 3 intraoperative rectal injuries (1 of these related to the Foley catheter placement noted above). The median number of hospital days during admission for RP was 2 (range, 1-12 days).
Endorectal MRI responses were evaluable in 38 patients. Reasons for the lack of endorectal MRI response data in the remaining 3 patients included discontinuation of treatment after cycle 3 (n = 1) and lack of measurable target lesions (n = 2). A waterfall plot detailing the post-treatment changes in tumor volume by endorectal MRI is shown in Figure 2. Twelve of 41 patients (29%; 95% CI, 16%-45%) achieved a >50% reduction in tumor volume.
A waterfall plot demonstrating the post-treatment changes in PSA is shown in Figure 2. Any degree of PSA decline was noted in 76% (95% CI, 60%-88%) of patients; 22% (95% CI, 11%-38%) achieved a >50% post-treatment decline in PSA. Several patients achieving a post-treatment decline in PSA experienced an increase in tumor volume on endorectal MRI and vice versa (Fig. 2).
Of the 37 patients who underwent RP, 24 (65%; 95% CI, 47%-80%) achieved a nadir <0.1 ng/mL postsurgery (Fig. 1). Nineteen of the 37 (51%) patients have not developed biochemical recurrence postprostatectomy (as defined by a PSA ≥0.2 ng/mL confirmed by subsequent PSA ≥0.2). The median time to recurrence from the date of prostatectomy was 13.0 months among the 37 patients who underwent RP (range, 1.2-41.5+ months). Among the 24 patients who had achieved a PSA nadir of <0.1 post-RP, the median time to biochemical recurrence from date of surgery had not yet been reached. The 2-year biochemical recurrence-free rate among the 24 patients was 68% (95% CI, 45%-83%).
The results of the central pathologic review are presented in Table 4. No patients achieved a pathologic complete response.
|Pathologic T stage|
The median testosterone level after completing chemotherapy was 382 ng/mL (range, 164-909 ng/mL). No patients experienced a decline in testosterone level to <50 ng/mL during treatment.
The current study demonstrates the safety and antitumor activity of neoadjuvant docetaxel plus bevacizumab in patients with high-risk localized PCa. Specifically, this study shows that this combination can be administered with rare grade >2 toxicities and uncommon intraoperative or perioperative complications. Furthermore, the majority of patients achieved some reduction in tumor volume by endorectal MRI and decreases in PSA. Unfortunately, similar to other reported neoadjuvant studies, no complete pathologic responses were observed. The rate of PSA progression in this high-risk cohort after treatment was higher than reported in some other neoadjuvant series,25 but similar to our prior study of neoadjuvant docetaxel alone.15 Although a worsening of outcomes with the addition of bevacizumab could be postulated, this is not highly suspected based on the available nonrandomized data.
The role of neoadjuvant docetaxel in high-risk localized PCa is the subject of ongoing randomized studies.16, 26 We previously demonstrated that single-agent docetaxel was associated with a reduction of tumor volume of >50% in 21% of patients,15 compared with 29% of patients in the current study of docetaxel plus bevacizumab. However, the small sample sizes of these studies and problems inherent in cross-study comparisons make conclusions regarding incremental improvements problematic. Furthermore, although tumor down-staging by MRI represents a noninvasive intermediate endpoint obtainable in the vast majority of patients on the current study, whether endorectal MRI response correlates with progression-free survival or overall survival has not yet been established and is the subject of ongoing work. Endorectal MRI may suffer from intraobserver variability, and other methodologies to aid in the interpretation of the results of single-arm neoadjuvant trials, such as comparing observed outcomes to nomogram-predicted outcomes, could also be considered. Also of note, in the current study, endorectal MRI responses did not directly correlate with post-treatment declines in PSA. Post-treatment decline in PSA may not be a useful endpoint in the neoadjuvant setting, a finding that warrants further evaluation, and correlation with long-term clinical outcomes, in an effort to define an optimal intermediate endpoint for neoadjuvant PCa studies. Although complete pathologic responses to neoadjuvant therapy have been associated with improved long-term outcomes in other solid tumors, complete pathologic responses have been an extremely rare event in neoadjuvant PCa studies, raising some concerns regarding the ultimate outcomes of ongoing randomized phase 3 trials evaluating neoadjuvant docetaxel in PCa. At the same time, the response rate to docetaxel in metastatic PCa is at least as high as the response rate to other single agents in solid tumors where chemotherapy has demonstrated benefit in the perioperative setting (eg, 5-fluorouracil in colon cancer), and relying only on changes in the primary tumor may not fully capture the potential benefits of systemic therapy on micrometastatic disease. Ultimately, the results of randomized trials will be required to determine whether the responses achieved with docetaxel-based neoadjuvant therapy in PCa are sufficient to improve clinical outcomes.
The results of the current study must be interpreted in the context of additional clinical data regarding the use of bevacizumab that have emerged since the conduct of this trial. CALGB 90401, in patients with castration-resistant PCa, demonstrated that although the combination regimen was associated with a statistically significant improvement in response rate and progression-free survival, there was no significant improvement in overall survival with the addition of bevacizumab to docetaxel.22 Despite these findings, the potential mechanism of action of antiangiogenic therapy in patients with bulky metastatic disease (eg, vascular normalization and improved chemotherapy delivery) may be quite different from the mechanism in patients at high risk for micrometastatic disease (eg, preventing metastases through inhibition of recruitment of the neovasculature). In this regard, the results of National Surgical Adjuvant Breast and Bowel Project C-08 and the AVANT study are more concerning. Both of these randomized phase 3 trials, exploring the use of adjuvant chemotherapy plus bevacizumab in high-risk localized colon cancer, failed to demonstrate an improvement in relapse-free survival with the addition of adjuvant antiangiogenic therapy.27 The results of ongoing randomized studies exploring perioperative bevacizumab in other solid tumors will be helpful in determining whether docetaxel plus bevacizumab warrants further evaluation as neoadjuvant therapy in PCa.
The role of integration of androgen deprivation therapy (ADT) into neoadjuvant regimens is also worthy of discussion. Several randomized trials have explored various durations of neoadjuvant ADT before prostatectomy.8-12 These trials have generally revealed improved tumor down-staging and a lower rate of positive surgical margins with the use of neoadjuvant hormonal therapy, but no significant impact on long-term clinical outcomes. PCa xenograft studies have demonstrated conflicting findings regarding the optimal treatment sequence on combination chemotherapy plus hormonal therapy.28, 29 Phase 2 studies have demonstrated the feasibility and activity of docetaxel plus ADT as neoadjuvant therapy before prostatectomy, with 1 study demonstrating 2 complete pathologic responses.13 Recently, preliminary results from the GETUG 12 trial, which randomized patients with high-risk localized PCa to neoadjuvant ADT alone versus ADT plus docetaxel and estramustine, followed by local therapy (radiation in the majority), were reported.26 The chemohormonal regimen was associated with a significant improvement in PSA response rate and a nonsignificant improvement in 4-year progression-free survival, although the event rate was lower than expected on both arms. The combined chemohormonal approach, followed by prostatectomy, is being further explored in CALGB 90203.16
In summary, the current study demonstrates the safety and activity of docetaxel plus bevacizumab as neoadjuvant therapy before prostatectomy in patients with high-risk localized PCa. Given that subsets of patients will likely benefit preferentially from both docetaxel and bevacizumab, molecular correlates of response to treatment on the current trial are being analyzed and will be reported separately. The role of neoadjuvant chemotherapy in PCa, and perioperative antiangiogenic therapy in solid tumors, requires further elucidation through the results of ongoing randomized clinical trials.
Supported by Genentech and Sanofi-Aventis.
CONFLICT OF INTEREST DISCLOSURES
P.F.: honoraria, Sanofi-Aventis. T.K.C.: consultant, Genentech. M.E.T.: consultant, Sanofi-Aventis. W.K.O.: honoraria, Sanofi-Aventis, Genentech.