|Home | About | Journals | Submit | Contact Us | Français|
This phase II trial was designed to evaluate the efficacy of vorinostat in chemotherapy pretreated patients with metastatic castrate resistant prostate cancer.
Patients with disease progression on one prior chemotherapy, a PSA ≥ 5ng/ml, and adequate organ function were treated with 400 mg vorinostat orally daily. The primary endpoint was the six month progression rate. Secondary endpoints included safety, rate of PSA decline, objective response, overall survival, and effects of vorinostat on serum IL-6 levels.
Twenty-seven eligible patients were accrued. Median number of cycles delivered was 2 (range 1–7). All patients were taken off therapy before six months. The best objective response in the eligible patients was stable disease in 2 (7%) patients. No PSA decline of ≥ 50% was observed. There was one grade 4 adverse event (AE) and 44% of patients experienced grade 3 AEs. Most common AEs were: fatigue (81%), nausea (74%), anorexia (59%), vomiting (33%), diarrhea (33%), and weight loss (26%). Median time to progression and overall survival were 2.8 and 11.7 months respectively. Median IL-6 levels (pg/ml) were higher in patients removed from protocol for toxicity vs. progression at all time points, including baseline (5.2 vs 2.1, p=0.02), Day 15-Cycle 1 (9.5 vs 2.2, p=0.01), Day 1-Cycle 2 (9.8 vs 2.2, p=0.01), and end of study (11.0 vs 2.9, p=0.09)
Vorinostat at this dose was associated with significant toxicities limiting efficacy assessment in this patient population. The significant association between IL-6 levels and removal from study for toxicities warrants further investigation.
With the establishment of docetaxel as standard first line chemotherapy for castrate resistant prostate cancer (CRPC) 1, 2, a clinical research priority in this disease is to identify second line therapy. Histone deacetylases (HDACs) regulate cell signaling and gene transcription through removal of acetyl groups from histone and non-histone proteins 3–5. Inhibition of HDAC activity leads to accumulation of acetylated proteins, which in turn lead to alterations in transcription, mitosis, and protein stability with resultant inhibition of tumor cell proliferation and survival 3–6. In preclinical studies, HDAC inhibitors have been shown to induce tumor cell cytostasis, differentiation, and apoptosis, and to inhibit tumor angiogenesis in various hematologic and solid malignancies, In prostate cancer, HDAC inhibition has resulted in decreased proliferation in cell lines7–9, and decreased tumor growth in preclinical models 9–15 suggesting that HDAC inhibition is of a potential therapeutic benefit in this disease.
Vorinostat is a small molecule inhibitor of class I and II HDACs that has been approved by the Food and Drug Administration for treatment of cutaneous T-cell lymphoma 16–18. In early testing, vorinostat showed significant antitumor activity in a broad range of cancers 19–22 including preclinical activity in prostate cancer 23, 24. Specifically, vorinostat suppressed the growth of the LNCaP, PC-3, and TSU-Pr1 cell lines at micromolar concentrations 23. In mice with transplanted CWR222 human prostate tumors, vorinostat treatment at 50 mg/kg/day resulted in significant suppression of tumor growth. At this dose, there was no detectable toxicity as evaluated by change in weight and necropsy examination 23. Kulp and colleagues have similarly shown growth inhibition of PC-3, DU-145, and LNCaP human prostate cancer cell lines and suppression of PC-3 xenograft tumors with vorinostat treatment 9. These biologic, preclinical and phase I data collectively provided the rational for testing vorinostat in patients with CRPC failing prior chemotherapy.
Interleukin-6 (IL-6) is a pleiotropic cytokine that stimulates the progression of a variety of cancers. Multiple studies have demonstrated that IL-6 is elevated in the sera of patients with metastatic prostate cancer 25–27. Drachenberg and colleagues 28 reported elevated serum IL-6 levels in men with hormone-refractory prostate cancer compared to normal controls, benign prostatic hyperplasia, prostatitis, and localized or recurrent disease suggesting that IL-6 may be a surrogate marker of the androgen independent phenotype. IL-6 has also been associated with disease progression and has been implicated as a potential marker of response to therapy 29–31. HDAC inhibition has also been shown to be associated with decreased expression of IL-6 and other pro-inflammatory mediators32–34 These findings, along with the observations that vorinostat can down-regulate the IL-6 signaling cascade 35 portends a possible role for the evaluation of IL-6 as an indicator of response to vorinostat. We hypothesized that vorinostat-mediated down regulation of IL-6 activity would be associated with a favorable outcome.
This Cancer Therapy Evaluation Program (CTEP) sponsored trial was conducted by the Department of Defense (DOD) Prostate Cancer Clinical Trials Consortium (PCCTC) and the National Cancer Institute (NCI)-sponsored University of Chicago Phase II Consortium. The protocol was reviewed and approved by the institutional review board at each participating institution and all patients provided informed consent prior to initiation of any study procedures. Eligible patients had metastatic prostate cancer with measurable and/or bony disease that had progressed despite androgen deprivation therapy and one prior chemotherapy regimen for CRPC. All patients were required to have prostate-specific antigen (PSA) progression defined as at least two rises in PSA documented over a reference value, no less than 7 days apart, with a minimum value of 5ng/ml. Patients had to have an Eastern Oncology Cooperative Group (ECOG) performance status of 0–2, adequate hematological, renal, and hepatic function defined by a white blood count of ≥ 3,000/μl, absolute neutrophil count ≥ 1,500/μl, platelet count ≥ 100,000/μl, creatinine < 2mg/dl, bilirubin within normal limits, aspartate aminotransferase (AST) and alanine transaminase (ALT) ≤ 2.5 X the upper limits of normal. Patients with significant cardiovascular disease including congestive heart failure (New York Heart Association Class III or IV), active angina pectoris or recent myocardial infarction (within the last 6 months) were excluded. Patients requiring diuretics for reasons other than hypertension, digoxin for reasons other than atrial fibrillation, or with a history of mild to moderate congestive heart failure, or patients with the following electrocardiogram (EKG) results: (a) significant q waves, (b) ST elevation or depressions of greater than 2 mm, (c) the absence of a regular sinus rhythm, or (d) the presence of a bundle block were required to undergo additional cardiac testing. Patients with known brain metastases were excluded but those with treated and controlled epidural disease were eligible. Patients on luteinizing hormone-releasing hormone (LHRH) agonists were required to continue therapy Discontinuation of all nonsteroidal antiandrogens (28 days for flutamide and 42 days for bicalutamide) was required. Patients taking valproic acid (a histone deacetylase inhibitor) must have stopped therapy at least two weeks prior to registration. No investigational or commercial agents (other than LHRH analogues) or therapies including other hormonal agents such as steroids, megesterol acetate (unless low dose given for hot flashes), antiandrogens or herbal medications were permitted to be administered with the intent to treat the patient’s malignancy. Patients with a “currently active” second malignancy other than non-melanoma skin cancers were not eligible. Patients were not considered to have a “currently active” malignancy if they had completed therapy and were considered by their physician to be with no evidence of disease.
Patients received open-label oral vorinostat 400 mg daily continuously. All patients completed a medication diary to monitor compliance. Toxicity was assessed using NCI-common terminology criteria (CTC) version 3.0 and dose reductions to 300 mg/day and 100mg/day were specified for grade 3 or 4 toxicities. Patients were evaluated clinically and by laboratory tests every 21 days. A maximum four week break in treatment for toxicity resolution was permitted.
Patients were monitored by history and physical exam, toxicity assessment, and PSA every three weeks. Response assessment by bone scan and CT scan and/or other appropriate imaging was performed every 12 weeks. Patients were removed from protocol if there was evidence of progression by PSA or Response Evaluation Criteria in Solid Tumors (RECIST) criteria, or symptomatic progression. Patients with bone scan only progression at first assessment continued treatment with reassessment after 6 additional weeks of therapy. Patients with confirmed progression were removed from protocol. Patients with stable disease or better were permitted to continue protocol therapy. Patients demonstrating progression by bone scan or other measures at the 24 week or subsequent scheduled assessments were considered as having progressive disease and a confirmation of progression was not required. All patients were followed for survival.
Progression for the purpose of the study was defined by any one or more of the following parameters: 1.) PSA Progression: 25% increase over baseline or nadir whichever is lower and an increase in the absolute value of PSA by 5ng/ml that is confirmed by another PSA at no less than a 4 week interval. 2.) Measurable disease progression: progression of target lesions by RECIST criteria 36 3.) Nonmeasurable disease progression: worsening of bone scan defined as development of ≥2 new lesions, appearance of new metastatic lesions outside of the bone, unequivocal progression of existing non-target lesions, or development of an indication for radiation therapy or other change in cancer therapy based on a change in a disease manifestation while on therapy.
Objective responses were defined using RECIST criteria 36. PSA “response” was defined based on the PSA Working Group Consensus Criteria 37. Bone disease was evaluated by bone scan with disease characterized as complete response if disappearance of all osseous lesions as evaluated by scans, stable or improved defined as no new lesions and no new pain in an area that uptake was previously visualized, and progression as defined by the appearance of two or more new skeletal lesions. An increase in the size or intensity of lesions was not considered progression.
The primary objective of this phase II trial was to evaluate the activity of oral vorinostat in patients with metastatic prostate cancer that had progressed on one prior chemotherapy regimen. The primary endpoint was the proportion of patients who did not demonstrate disease progression at 6 months. Based on a published retrospective analysis of second-line chemotherapy in men with metastatic CRPC 38, the expected progression rate by criteria used in this protocol in this patient population at six months is 84% (non progression rate of 16%). Therefore, if the progression-free rate was 10% or less, there would be little interest in pursuing this therapy further, whereas, with a progression-free rate of 30% or more, further study would be proposed.
Given the late time point for measuring progression, a single-stage design was used. Using Fisher’s exact test, 29 patients were to be accrued. If seven or more of these 29 patients were progression-free at six months, this agent would be felt to be worthy of further evaluation. This design provided for 80% power at the 5% significance level.
Secondary endpoints were to evaluate the safety of vorinostat and to determine the objective response rate in patients with bidimensionally measurable disease, the rate of PSA decline of ≥50%, progression-free and overall survival.
When designing this trial, we hypothesized that vorinostat-mediated down regulation of IL-6 activity would be associated with a favorable outcome. However, since all eligible patients were taken off study before six months, this analysis was not possible. Given that IL-6 is associated with the systemic immune response 39, we performed an exploratory analysis to determine if patients with higher levels of serum IL-6 were more likely to be removed from protocol for toxicity versus progression.
Pre-treatment and on-treatment peripheral blood samples for IL-6 were collected two hours following the most recent dose of vorinostat on day 15 of cycle 1, day 1 of cycle 2, the last week of cycle 4, and at time of removal from study. Quantitative levels of IL-6 were measured using a human IL-6 immunoassay (Quantikine® HS, R&D Systems) according to manufacturers instructions. IL-6 levels were compared between patients removed from protocol for progression vs. toxicity using the Wilcoxon rank-sum test.
Between 5/06 and 2/07, 29 patients were registered to protocol. Two patients were ineligible (due to noncastrate testosterone levels or previous treatment with a radiopharmaceutical). Table 1 lists baseline patient characteristics of the 27 eligible patients. The median age was 68 (range: 54–80). 70% of patients had a performance status of 1. Previous chemotherapy treatment for metastatic CRPC included docetaxel (92%), paclitaxel (4%), and cyclophosphamide (4%). All patients are off protocol therapy with a median number of cycles given of 2 (range 1–7). 70% of patients required dose reduction.
Forty-eight percent of patients experienced grade 3/4 toxicities. There were no grade-5 (treatment-related deaths) adverse events (AEs). Table 2 describes in detail toxicities by type and grade for which 70% of patients required dose reductions. The most common AEs were: fatigue (81%), nausea (74%), anorexia (59%), vomiting (33%), diarrhea (33%), and weight loss (26%). Eleven (41%) patients discontinued therapy due to toxicity (Table 3).
All eligible patients were off therapy before six months (Table 3); 13 (48%) were removed due to progression, 11 (41%) secondary to toxicity, and 3 (11%) for other reasons. The best objective response obtained was stable disease in two patients (7%). Duration of stable disease was 84 and 135 days, respectively. No PSA declines of ≥ 50% were observed (Figure 1). Median time to progression was 2.8 months (range 0.5–5.3) with a median overall survival of 11.7 months (2.3–14 months with one patient censored at 15.1 months). Of note, the two additional ineligible patients not included in the final analysis also achieved a best stable response of stable disease.
Median IL-6 levels (pg/ml) were higher in patients removed from protocol for toxicity vs. progression at all time points, including baseline (5.2 vs 2.1, p=0.02), Day 15-Cycle 1 (9.5 vs 2.2, p=0.01), Day 1-Cycle 2 (9.8 vs 2.2, p=0.01), and end of study (11.0 vs 2.9, p=0.09) (Figure 2).
To date there is no established second line systemic therapy for patients with CRPC. HDAC inhibitors are attractive agents, particularly in prostate cancer, due to demonstrated effect in vitro on proliferation, differentiation, apoptosis, and angiogenesis coupled with anti-tumor effects in preclinical prostate cancer models.
Recognizing that tumor regressions are difficult to quantify objectively in patients with bone metastases, the clinical importance of delaying progression, and the available preclinical data on the anti-tumor effect of vorinostat, this trial was designed with a primary objective of assessing the effect of vorinostat on six month progression rates. Although, the most optimal design would have included a control arm, the progressive nature of this disease and the availability of published historical institutional data, at time of study design, on second-line chemotherapy in a similar population indicating that the expected 6-month progression rate is 84% 38 lead us to chose a single arm design. Although 41% of patients were taken off study due to toxicity thus making it difficult to assess the true efficacy of vorinostat at this dose and schedule, it is reasonable to assume that, had there been clinically meaningful anti-tumor activity, better results would have been expected. There was only one grade 4 AE, and grade 3 AE’s were predominantly constitutional in nature and not significantly different from dose limiting toxicities observed in phase I testing,21. However, despite dose reduction in 70% of patients in this trial, 41% of patients discontinued therapy due to toxicity. Our experience is in contrast with other reports using this agent both as monotherapy and in combination with other systemic therapies in other studies. In the phase I trials, once on a tolerable dose, patients could be treated for prolonged periods of time 21, 22, 40 with chronic adverse effects of fatigue, renal insufficiency, and weight loss reversible upon discontinuation of drug.21 Dose limiting toxicities reported in phase I trials were not related to prior therapy or type of underlying malignancy and remained unpredictable within treatment cohorts.21 They were also rapidly reversible suggesting a readily reversible metabolic process.21
Safety data from 86 patients with cutaneous T-cell lymphoma treated with vorinostat leading to FDA approval of the drug, only 9.3% of patients were removed due to toxicity with 10.5% requiring dose reductions using the same dose/schedule as used in this trial; also in patients who had failed prior systemic therapies.16 Similar results were recently reported on safety data from 476 patients who participated in the vorinostat clinical trial program receiving vorinostat as single agent therapy and in combination with other systemic therapies 41.
The key question is whether our observed results are a function of the patient population studied or lack of significant anti-tumor activity or both. Given the toxicity seen in this trial leading to dose reductions in 70% of patients, it is possible that suboptimal cell inhibitory plasma concentrations of vorinostat may explain why less clinical activity was seen than expected. Without pharmacokinetic (PK) data and data from other prostate cancer settings, it is difficult to conclude whether the preclinical models were poor predictors of clinical activity or whether this agent would be more efficacious in an alternative patient population or dosing schedule. One interesting observation from this population is that patients that came off study due to toxicity had significantly higher serum IL-6 levels at all time points (baseline, Day 15-Cycle 1, Day 1-Cycle, and end of study) as compared to patients removed from study for progression. It is possible that since IL-6 is associated with the inflammatory response and regulation of the systemic immune response 39, that higher levels of serum IL-6 at baseline that were not modulated by the drug, predisposed patients to adverse side-effects leading to treatment discontinuation. IL-6 has been associated with non-responsiveness to drug therapy 29–31. However, of the 11 patients taken off protocol due to toxicity in this study, nine patients recovered suggesting drug effect and not disease progression.
Toxicities were also prominent with no significant clinical activity in the only other reported clinical trial of HDAC inhibition in prostate cancer 42. In this phase II trial (n=31) investigating romidepsin, a bicyclic depsipeptide that inhibits HDAC, as front line therapy for patients with metastatic CRPC, constitutional toxicities were common, with a 6 month disease control rate of 14% and PSA response rate of 7%. Observations from this and our trial raise questions regarding the impact of an androgen suppressed state as it relates to predisposing to toxicities to this class of drugs.
It is not clear why outcomes from clinical investigation of HDAC inhibitors in metastatic CRPC have not matched the promising preclinical activity and scientific rationale. However, based on the current data, further investigation of vorinostat at this dose and schedule is not recommended. Given the lack of significant clinical activity in this trial coupled with comparable outcome reported with romidepsin 42 raise concerns regarding further study of this class of drugs as single agent therapy for treatment of CRPC unless newer agents with a more favorable toxicity profile with substantial supportive preclinical data are introduced. Our observation of the association of IL-6 levels and removal from study for toxicities warrants further investigation.
CTEP, PC051382, PC051375, N01-CM-62201, PCF N008367 D.A.B is supported by NIH grant T32CA009357, UL1RR024986.
Authors of the DOD Prostate Cancer Clinical Trial Consortium and Phase II consortium also included George Wilding4, Susan Slovin2, Kathleen A. Cooney1, June Escara-Wilke1and Evan Keller1.
This study was presented in part at the 2008 ASCO Genitourinary Cancers Symposium