Randomised phase III study of intravenous vinorelbine plus hormone therapy versus hormone therapy alone in hormone-refractory prostate cancer

R. P. Abratt1,*, D. Brune2, M.-A. Dimopoulos3, J. Kliment4, J. Breza5, F. P. Selvaggi6, P. Beuzeboc7, T. Demkow8 and S. Oudard9

1 University of Cape Town, South Africa; 2 Centre François Baclesse, Caen, France; 3 Alexandra Hospital, Athens, Greece; 4 Je Senius School of Medicine, Martin; 5 Comenius University Medical School, Bratislava, Slovak Republic; 6 Ospedale Policlinico Consorziale, Bari, Italy; 7 Institut Curie, Paris, France; 8 Maria Sklodowska Curie Memorial Cancer Center, Warsaw, Poland; 9 Hopital Européen Georges Pompidou, Paris, France

* Correspondence to: Dr R. P. Abratt, Department of Radiation Oncology, Groote Schuur Hospital, Observatory, 7925, Cape Town, South Africa. Tel: +27-21-404-4266; Fax: +27-21-448-5707; Email: rpa{at}curie.uct.ac.za


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background: Vinorelbine (VRL) has been shown to be active in hormone-refractory prostate cancer (HRPC) in phase II studies, alone or in combination. Its moderate toxicity profile is well tolerated in elderly patients.

Patients and methods: Patients with metastatic prostate cancer, progressive after primary hormonal therapy, were randomised to receive intravenous VRL 30 mg/m2 on days 1 and 8 every 3 weeks, and hydrocortisone 40 mg/day or hydrocortisone alone until disease progression. Centres could choose to add aminoglutethimide 1000 mg/day to hydrocortisone as second-line hormone therapy (HT) for all their patients. Randomisation was stratified by centre. Further chemotherapy was allowed after progression. The primary end point was progression-free survival (PFS). The final analysis was performed on a total of 414 patients. Reported results were all based on intention-to-treat analyses. All progressions and responses were reviewed by an independent panel.

Results: PFS was significantly prolonged in the VRL plus HT arm compared with the HT alone arm, according to the statistical hypothesis of the protocol (P=0.055 in the two-sided log-rank test with a pre-specified significance level of 10%). The 6-month PFS rates were 33.2% versus 22.8%, and the median durations of PFS were 3.7 versus 2.8 months. In the multivariate Cox analysis, which included age, Karnofsky performance status (PS), haemoglobin, alkaline phosphatase at study entry and number of prior hormonal treatments, the P value was decreased to 0.005. The prostate-specific antigen (PSA) response rate (≥50% decline sustained for at least 6 weeks) was significantly higher for VRL plus HT compared with HT (30.1% versus 19.2%; P=0.01). Clinical benefit, defined as a decrease in pain intensity or analgesic consumption or an improvement of Karnofsky PS for at least 9 weeks, and at least stable assessment in the other two, was also more frequently observed in patients who received VRL plus HT versus HT alone (30.6% and 19.2%; P=0.008). There was no statistical difference in overall survival. Forty-three per cent of patients in the HT arm received at least one line of further chemotherapy after progression, compared with 28% of patients in the VRL-based arm. Aminoglutethimide did not seem to result in better efficacy for either arm. VRL plus HT was well tolerated, with a median administered relative dose intensity of 90%; grade 4 neutropenia occurred in 6.5% of patients and non-haematological toxicity was rare.

Conclusions: The combination of VRL and hydrocortisone compared with hydrocortisone alone resulted in improved clinical benefit, PFS and PSA response rate. This therapeutic gain is similar to that previously reported with mitoxantrone in combination with low-dose corticosteroids. There was no gain in survival; however, the combination is well tolerated in this elderly group of patients, who often present cardiac co-morbidities, and therefore offers an active and safe therapeutic option for patients with hormone-refractory prostate cancer.

Key words: hormone-refractory prostate cancer, hormone therapy, randomised phase III clinical trial, vinorelbine


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Androgen deprivation in patients with metastatic prostate cancer produces effective palliation and normalisation of abnormal prostate-specific antigen (PSA) in the majority of patients, but all patients will progress and eventually become hormone-refractory. Following disease progression after primary hormone therapy (HT), some remaining androgen-sensitive cells may respond to secondary hormonal manipulation [1Go]. Second-line HT, which consists of anti-androgen withdrawal, corticosteroids, oestrogens, aminoglutethimide or ketoconazole, produces symptomatic improvement in 10% to 20% of hormone-refractory prostate cancer (HRPC) patients [2Go], but the duration of response is usually short. The role of chemotherapy has therefore been evaluated and palliative benefit has been reported in phase III studies [3Go, 4Go].

Vinorelbine (VRL) is a semi-synthetic vinca alkaloid that has undergone phase II studies in HRPC. Single-agent studies showed a ≥50% decrease in PSA levels sustained for 3–4 weeks in 13% to 17% of evaluable patients [5Go–7Go], and durable clinical benefit as defined by improvement in pain index and/or performance status (PS) in 32% to 39% of patients [5Go, 6Go]. Toxicity was moderate and was mainly haematological with grade 3–4 neutropenia reported in 51% of patients receiving VRL at 25 mg/m2/week [6Go] and 20% of those receiving the drug on days 1 and 8 of a 3-week cycle [7Go]. VRL is well-tolerated in elderly patients, as shown in studies of non-small-cell lung cancer (NSCLC) and breast cancer [8Go–10Go], which makes it an attractive candidate in HRPC patients, whose median age is 70 years. A modified dosage regimen of 30 mg/m2 on days 1 and 8 of a 3-week cycle was investigated exclusively in elderly patients with NSCLC and generated an overall decrease in haematotoxicity while significantly improving survival over best supportive care [8Go].

The primary objective of this phase III trial was to determine whether the combination of intravenous VRL in combination with second-line HT increased the progression-free survival (PFS) of HRPC patients in comparison with second-line HT alone. Second-line HT was daily hydrocortisone with or without aminoglutethimide depending on the individual centre's decision at study initiation. Secondary objectives included the comparison of PSA response rates, clinical benefit response rates, tumour response rates and overall survival.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Between April 1997 and August 2001, 451 patients entered this trial from 50 institutions. The study was approved by the respective national health authorities and the ethics committees of the participating institutions. An independent review in December 2000 detected major data inconsistencies and irregularities in all 37 patients from one country. These findings were confirmed by a for-cause audit performed on-site. All these 37 patients were excluded from the analysis, which was statistically acceptable as patients were randomised by centre. Therefore, a total of 414 patients were included in the final study analysis.

Eligibility
Patients with histologically documented, metastatic adenocarcinoma of the prostate who had failed prior androgen deprivation and with castrate level of testosterone were considered eligible for the trial. Progressive disease was required for study entry, and was defined by either objective evidence of disease progression (appearance of a new measurable or evaluable lesion, or increase in any previously existing measurable lesion ≥25% in cross-sectional area, or appearance of a new bone lesion or de novo urinary obstruction), or by progressive symptoms with rising PSA in a patient with metastatic disease. PSA value at study entry had to be >10 ng/ml. Between 10 and 20 ng/ml, PSA values must have doubled compared with values before progression. For upper values >20 ng/ml, it was not necessary to wait for doubling of PSA value. Progressive symptoms were defined by either a deterioration of PS by at least one general category, a cancer-related weight loss of at least 10% or an increased cancer-related pain. No pain stabilisation was required before study entry. Patients previously treated with an anti-androgen were required to have progressive disease despite anti-androgen withdrawal. A minimum of 4 weeks was required to have passed for flutamide and nilutamide and 8 weeks for bicalutamide. Other criteria were an age ≥18 and ≤85 years, a Karnofsky PS ≥60%, adequate organ function defined as a neutrophil count ≥2 x 109/l, haemoglobin ≥9 g/dl, platelet count ≥100 x 109/l, total bilirubin ≤1.5x the institutional upper limit of normal (ULN), transaminases ≤2.5x ULN and creatinine ≤1.5x ULN. Radiotherapy had to be completed at least 2 weeks before study entry and strontium-89 therapy completed at least 8 weeks before study entry. Patients not eligible for study entry were those who had received prior chemotherapy including estramustine phosphate; presented with symptoms suggesting brain metastases; were medically unstable; had uncontrolled infection at study entry; received concurrent treatment with biphosphonates or any experimental drug; or who had a past history of another malignancy except basal or squamous cell carcinoma of the skin. Patients with cardiac co-morbidity were eligible for entry into the study. All patients were required to sign an informed consent before study entry, according to local regulatory requirements.

Randomisation and treatment
Randomisation was stratified by centre only. Patients were randomised to receive either VRL 30 mg/m2 on days 1 and 8 of a 3-week cycle plus hydrocortisone 40 mg/day, or hydrocortisone 40 mg/day. Individual centres could choose to add aminoglutethimide 1000 mg/day to either study arm; they were to select the type of second-line HT (hydrocortisone alone or hydrocortisone plus aminoglutethimide) to be used in both study arms before enrolment of the first patient, and were not to deviate from this regimen throughout the trial.

Patients who had not undergone orchiectomy had to continue luteinising hormone-releasing hormone (LH-RH) agonist therapy while on study.

Treatment was given for at least 9 weeks (three cycles) unless early disease progression (defined as a progression occurring before the first tumour assessment at 9 weeks) or unacceptable toxicity occurred. At the end of this three-cycle period, patients with a response or stable disease continued the treatment until disease progression, excessive toxicity or patient refusal.

VRL was given if neutrophils count was >1.5 x 109/l and platelet count >75 x 109/l. If not, the day 8 administration was omitted and the day 1 administration was delayed for 1 week (for a maximum of 2 weeks). No dose reduction schema for VRL was allowed.

Patients in the second-line HT alone arm were permitted to receive chemotherapy including VRL after disease progression.

Evaluation
Evaluation before the start of treatment included medical history, physical examination, Karnofsky PS, pain intensity, analgesic consumption, complete blood cell count, serum chemistry, measurement of serum PSA, testosterone level, chest X-ray, abdomino-pelvic computed tomography scan, bone scintigraphy, electrocardiogram and quality of life evaluation. During treatment, Karnofsky PS, pain intensity, analgesic consumption, blood cell counts and PSA level were determined every 3 weeks. Blood cell counts were repeated on day 8 of each cycle in patients treated with VRL. Serum chemistry, bone scintigraphy, other radiological tests needed to evaluate tumour lesions and quality of life were assessed every 9 weeks.

The study protocol used the following criteria to evaluate the effect of treatment on PSA. A complete response (CR) required a PSA normalisation (defined as <0.2 ng/ml) confirmed by two further determinations over a 6-week period; a partial response (PR) was defined as a ≥50% decline compared with baseline confirmed by two further determinations for a minimum of 6 weeks [11Go]; disease progression was defined as two consecutive increases ≥100% compared with baseline or nadir; stabilisation was recorded when the criteria for PR or progression were not met.

A clinical benefit response was defined by a positive response in one of three categories assessed (pain intensity, analgesic consumption or Karnofsky PS) for at least 9 weeks, and at least a stable assessment in the other two categories. Patients were considered stable for clinical benefit evaluation if the three categories assessed remained stable for at least 9 weeks. This concept of clinical benefit measured by two common debilitating signs or symptoms in patients with advanced prostate cancer has already been used and validated [5Go]; this end point was derived from a palliative end point used for pancreatic cancer [12Go]. Pain intensity was assessed on a four-point discontinuous scale on which 0 = no pain and 3 = severe pain. Analgesic consumption was assessed on a five-point scale on which 0 = no requirement for analgesics, 1 = occasional non-opioid analgesics, 2 = regular non-opioid analgesics, 3 = regular weak opioids and 4 = regular morphinic analgesics. For pain intensity and analgesic consumption, a positive response was a decrease of one category. For Karnofsky PS, a positive response was an increase of 10%.

Tumour response was evaluated by WHO criteria [13Go].

The primary study end point was PFS, which was defined as the time elapsed from the date of randomisation until progression or death (whatever the cause) or last news or initiation of a new therapy. In the last two instances, the patient was censored. Progression was documented on the first of the three end points that met the criteria of progression as follows:

  1. appearance of a new lesion or a ≥25% increase in an existing lesion;
  2. PSA progression as defined above; or
  3. an increase by one category of pain intensity and an increase by one category of analgesic consumption; or a deterioration of 20% on the Karnofsky PS when it was in the 70% to 100% range and a deterioration of 10% when it was 60%, each maintained on two consecutive evaluations. Progression was determined in comparison with the smallest score reported for pain and analgesic consumption and the highest score for PS.

Quality of life was evaluated using the European Organisation for Research and Treatment of Cancer (EORTC) QLQ-C30 questionnaire.

Toxicity was graded according to the Cancer and Leukemia Group B expanded common toxicity criteria. Febrile neutropenia was defined as an absolute neutrophil count <0.5 x 109/l concomitant with fever ≥38.5°C. Neutropenic infection was defined as grade ≥3 infection concomitant with absolute neutrophil count <1 x 109/l.

Independent review committee
The three efficacy parameters from the entire study population (451 patients) and dates of progression were reviewed by an independent review committee (IRC) consisting of two independent oncologists. The clinical and radiological review was blinded to treatment allocation.

Statistical methods
The planned sample size of 362 patients was based on the hypothesis of demonstrating a 30% increase of PFS, which was clinically relevant, with an {alpha} of 10% and a ß of 20%.

All analyses were performed on an intention-to-treat basis, and based on data validated by the IRC. The primary analysis was the comparison of PFS between VRL plus second-line HT and second-line HT alone, which was performed with a stratified-by-centre log-rank test. This trial was designed to have a 80% power at a 10% two-sided significance level. In order to respect the usual 5% level for {alpha} risk, 95% confidence intervals (CIs) were calculated for each efficacy parameter that was analysed. To describe time-dependent parameters, the Kaplan–Meier product limit method was used. Multivariate analysis was performed using a Cox model adjusted for age, Karnofsky PS and PSA level at baseline. In addition to those prognosis factors identified in the study protocol, other well-established factors [14Go, 15Go], including number of prior hormonal manipulations, non-respect of the anti-androgen withdrawal period, haemoglobin value at baseline and alkaline phosphatase value at baseline, were also considered.

The {chi}2-test was performed to compare proportions; 95% CIs with normal approximation were provided.

In order to estimate agreement between PSA response and clinical benefit response at 9 weeks, the Kappa coefficient and statistical test were used.

A subgroup analysis on the impact of aminoglutethimide could be performed because the randomisation was stratified by centre and accordingly by treatment with aminoglutethimide.

An intention-to-treat analysis based on investigators' data has also been performed.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Among the 414 patients included in the final study analysis, 206 were randomised to receive VRL plus second-line HT and 208 to second-line HT alone. A total of 37 patients (8.9%), equally distributed in the two study arms, were found to be ineligible. Reasons for ineligibity were absence of metastatic disease (n=13), discontinuation of LH-RH agonist therapy at study entry (n=9), no prior medical or surgical castration (n=6), prior treatment with estramustine (n=4), PSA level at study entry <10 ng/ml (n=3), prior or current history of another malignancy (n=2), concomitant biphosphonates (n=1), suspicion of concomitant lung cancer (n=1) and anti-androgen started just before inclusion (n=1). Some patients were ineligible for multiple reasons. These 37 patients were included in all the analyses of efficacy parameters.

Patient and tumour characteristics
Baseline characteristics are listed in Table 1. The patient characteristics were well balanced. However, there was a trend for patients randomised to receive VRL plus HT to have higher median PSA and alkaline phosphatase values at baseline. In addition, 22% of the patients in the VRL plus HT arm and 17% in the HT alone arm were enrolled after more than one prior hormonal manipulation excluding anti-androgen withdrawal. Anti-androgen therapy had been previously received by 72% of the study population. Among those patients, the non-steroidal anti-androgen withdrawal period was not respected for 11% of the patients from the VRL plus HT arm and 13% from the HT alone arm.


View this table:
[in this window]
[in a new window]
 
Table 1. Patients characteristics at study entry by treatment arm

 
Progression-free survival
PFS was the primary end point of the study. At the time of data analysis, the number of events was 352 (175 in the VRL arm and 177 in the HT arm). The intention-to-treat analysis of PFS showed a significant prolongation in the VRL plus HT arm compared with the HT alone arm according to the statistical hypothesis of the protocol (P=0.055 in the stratified-by-centre log-rank test with a pre-specified significant level of 10%; Figure 1). The median PFS was 3.7 months (95% CI 2.8–4.3) versus 2.8 months (95% CI 2.2–3.5) and the 6-month PFS rates were 33.2% and 22.8%, respectively. The criteria on which disease progression was documented are given in Table 2. In both study arms the mode of progression was almost equally divided between tumour progression (most often appearance of a new bone lesion), rising PSA and worsening symptoms such as pain or deterioration of PS.



View larger version (11K):
[in this window]
[in a new window]
 
Figure 1. Progression-free survival. VRL, vinorelbine; HT, hormone therapy.

 

View this table:
[in this window]
[in a new window]
 
Table 2. Reasons for progression

 
Multivariate analysis using a Cox model evaluated the impact of predetermined prognosis factors, which included age (>65 versus ≤65 years), PSA level at baseline (≥ median value versus < median value) and Karnofsky PS (≥80% versus <80%). In addition, baseline haemoglobin level (≥13 versus <13 g/dl), baseline alkaline phosphatase level (>2x versus ≤2x ULN), number of prior hormonal manipulation (1 versus >1) and respect of the anti-androgen withdrawal period (yes versus no) were incorporated in the model.

Results of the multivariate analysis after adding the treatment arm are displayed in Table 3. This analysis showed a decrease by 29% of the progression risk in patients treated with VRL plus HT compared with HT alone (P=0.005). Factors associated with a shorter PFS were low value of baseline haemoglobin, high value of baseline alkaline phosphatase, poor PS at study entry and more than one prior hormonal manipulation. Although baseline PSA value was significant in the univariate analysis, it proved to be non-significant in the multivariate analysis.


View this table:
[in this window]
[in a new window]
 
Table 3. Cox model for progression-free survival

 
PSA evaluation
PSA response rate (≥50% decline of PSA sustained for at least 6 weeks) was 30.1% (95% CI 24% to 36%) in the VRL plus HT arm and was 19.2% (95% CI 14% to 25%) in the HT alone arm. This difference between the two arms was significant (P=0.01). A PSA decrease ≥80% was also determined in the VRL plus HT arm (15%) and in the HT alone arm (7.2%). In each arm, six patients (2.9%) showed a normalisation of their PSA under 0.2 ng/ml.

Clinical benefit
The clinical benefit response rate in the VRL plus HT arm was 30.6% (95% CI 24% to 37%) and in the HT alone arm was 19.2% (95% CI 14% to 25%). This difference was statistically significant (P=0.008)

The distribution of responses for clinical benefit among patients who did or did not achieve a ≥50% decline in PSA after having received three cycles of study treatment is shown in Table 4. Agreement between PSA response and clinical benefit response at 9 weeks was moderate according to the Laudis and Koch scale [16Go]: the Kappa coefficient was 0.46 in the VRL arm and 0.41 in the HT alone arm. However, PSA and clinical benefit assessment were statistically correlated at 9 weeks (P<0.001) in both arms.


View this table:
[in this window]
[in a new window]
 
Table 4. Responses for clinical benefit according to PSA response at 9 weeks

 
Tumour response
A total of 142 patients had measurable metastases, 68 in the VRL plus HT arm and 74 in the HT alone arm. Four PRs were documented in patients treated with VRL (5.9%; 95% CI 0.3% to 11.5%) and none in those treated with HT.

Overall survival
Median follow-up was 24 months (range 4.5–57) in the VRL-based arm and 24.6 months (range 4.5–56.3) in the HT alone arm. Median survival was similar in both arms: 14.7 months (95% CI 11.8–18) in the VRL plus HT arm versus 15.2 months (95% CI 12.2–17.2) in the HT alone arm. About the same percentage of patients were still alive at 6 and 12 months in both arms (79.5% and 56.1% for VRL plus HT and 78.9% and 57.7% for HT alone, respectively). At 18 months, 41.1% of patients treated with VRL were still alive compared with 40.1% of those treated with HT alone. As shown in Figure 2, the survival curves are superimposable.



View larger version (17K):
[in this window]
[in a new window]
 
Figure 2. Overall survival. VRL, vinorelbine; HT, hormone therapy.

 
A greater number of patients in the HT alone arm received second-line chemotherapy compared with those in the VRL plus HT arm [89 (42.8%) and 58 (28.2%), respectively]. The predominant cytotoxic agent used was estramustine phosphate, in about half of retreated patients for both arms. In the HT alone arm, 8.2% of patients received second-line VRL. The proportion of patients who received a further line of HT after study treatment discontinuation was very similar in both arms (14.9% in the HT alone and 11.7% in the VRL plus HT arm).

Efficacy results based on investigator assessment
The efficacy analyses based on investigators' data (i.e. data collected in the case report form) are summarised in Table 5. The median PFS was 4.49 months in the test arm (95% CI 4.11–5.16) versus 3.13 months in the control arm (95% CI 2.66–3.91). The difference between the two arms is significant (P=0.007).


View this table:
[in this window]
[in a new window]
 
Table 5. Efficacy results based on investigator assessment

 
In an effort to explain the discrepancies between the results reported by the investigators and those by the IRC, reasons for progression were compared. The investigators reported progression on at least two criteria more frequently: in 19.8% of patients versus 2.5% as per IRC in the test arm, and 24.4% versus 7.3% in the control arm. Investigators usually did not take into account a progression based only on clinical benefit, but generally waited for objective documentation of progression on tumour lesion and/or PSA before considering the disease as progressive.

By comparison with IRC, the investigators tended to underestimate the PSA response and to overestimate the tumour response in both study arms. Assessment of clinical benefit as derived from the reporting of pain score, analgesic consumption and PS by the investigators was not markedly modified by the IRC.

Impact of aminoglutethimide on the primary end point
PFS was compared in the subgroups of patients who did or did not receive aminoglutethimide in each study arm. Aminoglutethimide was given in 180 patients (43%) of the study population. In patients treated with VRL plus HT, the median PFS in the subgroup receiving aminoglutethimide was 3.4 months, and in the subgroup not receiving aminoglutethimide was 4.2 months. In patients treated with HT alone, median PFS in the subgroup receiving aminoglutethimide was 2.6 months, and in patients not receiving aminoglutethimide was 3 months. There was no difference in the PFS for patients who did or did not receive aminoglutethimide in either the VRL (P=0.19) or HT alone arm (P=0.17) by the log-rank test.

Quality of life
Quality of life was assessed by the EORTC QLQ-C30 instrument, which is a general rather than a prostate cancer-specific tool. Of the 163 patients who completed three cycles of treatment, questionnaires were completed before randomisation by 62% of patients in both arms, and thereafter by 62% in the VRL plus HT arm and 64% in the HT alone arm. There were therefore difficulties in obtaining compliance with quality of life evaluation in this population of patients, and the findings need to be interpreted with caution. The mean global score of quality showed no difference between the two arms at baseline and at the first and second evaluations at 9 and 18 weeks (see Figure 3). There appeared to be a modest improvement in both arms at follow-up, although this might be an artefact owing to the loss of information from the more ill patients.



View larger version (11K):
[in this window]
[in a new window]
 
Figure 3. Gobal quality of life; VRL, vinorelbine.

 
Concerning the functional scale (including quality of life, physical, role, emotional, cognitive and social subscales) and the symptom scales (including evaluation of fatigue, nausea, pain, dyspnoea, sleep disturbance, appetite loss, constipation, diarrhoea and financial impact), the only significant differences noticed were a time effect in pain (P<0.0001), fatigue (P=0.0018) and nausea/vomiting (P=0.0253). The mean scores of these three parameters decreased in both arms; however, there were no significant differences between the two treatment arms.

Treatment delivery and toxicity
Patients tended to receive a higher number of cycles in the VRL arm than in the HT arm, the median number of cycles being six and four, respectively. There was a delay in the administration of chemotherapy in 16% of cycles and an omission of the day 8 administration of VRL in 6% of cycles. In the VRL arm, the majority of patients could receive the scheduled regimen of intravenous VRL and the median relative dose intensity was 90%.

As expected from the toxicity profile of VRL, the main toxicity in the VRL plus HT arm was neutropenia, with 26% of patients experiencing a grade 3 or 4 event. However, febrile neutropenia occurred in only one patient, and neutropenic infection was seen in six patients (3%). Severe anaemia and thrombocytopenia were rare and their incidences were similar in the two study arms.

Grade 3–4 non haematological toxicities were uncommon, as shown in Table 6. Infection including neutropenic infection was more frequently reported in patients treated with VRL (5%) than in those treated with HT alone (0.5%). Severe neuroconstipation was rare, and was only seen in patients treated with VRL (3%). The difference between the two arms was statistically significant only for leucopenia (P<0.001), neutropenia (P<0.001), infection (P=0.005) and neuroconstipation (P=0.014).


View this table:
[in this window]
[in a new window]
 
Table 6. Percentage of patients with specific toxicities

 
While 49% of the patients treated with VRL had cardiac co-morbidity consisting of high blood pressure (30% of patients), ischaemic cardiopathy (12%), dysrhythmia (7%), arteritis (5%) and thromboembolic disease (4%), very few cardiac disorders were reported. Two patients experienced transient grade 1 cardiac dysrhythmia in the VRL plus HT arm and one grade 1 episode of cardiac ischaemia and one grade 1 episode of cardiac dysfunction were observed in the HT alone arm.

Aminoglutethimide had no major impact on the safety profile. Overall, four patients were withdrawn from study treatment due to the occurrence of skin allergy (n=2) and neurocortical toxicity (n=2).


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
The aim of chemotherapy in HRPC is to prolong patients' survival and improve their well-being. Other indicators of antitumour activity are the tumour response to therapy and PFS. In our study, the clinical benefit response rate was higher in the VRL and HT arm than in the HT alone arm. This reflected improvement in pain and PS for a minimum of 9 weeks. Pain improvement is an important component in the treatment of patients with HRPC. The VRL arm was well tolerated, with a low side-effect profile. There was no difference in overall survival between the two arms; however, PFS was the primary study end point, and was statistically significantly longer in the test arm. The PSA response rate was also significantly higher in the VRL arm.

Results reported for hydrocortisone alone in the control arm are consistent with published experience in HRPC [3Go, 17Go, 18Go]. Clinical testing with aminoglutethimide was conducted in combination with hydrocortisone [19Go, 20Go]. However, in both treatment arms of the present study, the tendency was away from an increased PFS with aminoglutethimide, and therefore the use of aminoglutethimide in combination with VRL and hydrocortisone cannot be recommended.

The therapeutic advantage shown in the present study is similar to the one obtained with a mitoxantrone and low-dose corticosteroids combination in two previous phase III studies [3Go, 17Go]. However, patients' selection criteria differed from the present study: cardiac compromise was an exclusion factor for study entry in the mitoxantrone studies, and median time from diagnosis tended to be longer. In addition, the present study used more stringent criteria of disease progression, which included tumour, PSA, pain, analgesic consumption and PS assessments.

Indeed, taxanes seemed to show a greater clinical activity with PSA response rates and clinical benefit response rates up to 70% in combination with estramustine in several phase I and II studies [21Go, 22Go]. There was also a trend towards improved survival, with time to progression varying from 5 to 6 months and a median survival of 20–23 months [22Go–24Go]. These results have very recently been confirmed by two large phase III studies, showing for the first time a significant survival benefit in HRPC, in favour of docetaxel-based arms [25Go, 26Go]. The primary end point for both studies was overall survival. The Southwest Oncology Group 99-16 trial compared the current standard of care, mitoxantrone combined with prednisone, with docetaxel 60 mg/m2 every 3 weeks combined with estramustine 280 mg three times a day, days 1–5; the results showed a median survival of 18 months in the docetaxel plus estramustine arm versus 16 months in the mitoxantrone plus prednisone arm (P=0.01). Median PFS was 6 months versus 3 months (P<0.0001), and the PSA response rate was 50% versus 27% (P>0.0001) in the control arm [25Go]. The TAX 327 study compared docetaxel 30 mg/m2 given weekly or 75 mg/m2 every 3 weeks along with prednisone, with mitoxantrone plus prednisone in patients with HRPC; the results showed that only in the docetaxel 3-weekly arm was there a significant improvement in survival, with a median of 18.9 months versus 16.5 months in the mitoxantrone arm (P=0.009). Forty-five per cent PSA response rate was observed in the docetaxel 3-weekly arm versus 32% in the mitoxantrone arm (P=0.0005), and pain response rate was 35% versus 22% (P=0.01) [26Go]. In both studies, the toxicity profile was mild. These results set up a new standard of care for patients with prostate cancer who have failed HT, and will probably change the perception of chemotherapy in the field of prostate cancer.

VRL in combination with HT was well tolerated in this elderly population. The main dose-limiting toxicity was neutropenia, which was rarely complicated. The median relative dose intensity with the day 1 and 8 schedule was 90%, which is consistent with clinical experience in NSCLC patients. Non-haematological toxicity was minimal, the most prominent events being occasional episodes of mild and reversible neuroconstipation and peripheral neuropathy. It should be noted that compared with the two docetaxel phase III trials, and also the two mitoxantrone phase III trials, patient selection criteria were less restrictive in the VRL study regarding associated morbidities, especially cardiac: patients were excluded from the docetaxel and mitoxantrone studies, but not from the VRL study, if they had prior congestive heart failure, even if controlled, and if they had a previous history of myocardial infarction or angina pectoris within 1 year, or uncontrolled hypertension or dysrythmia. Cardiac toxicity was remarkably low (<1% of patients) in the present trial, despite the fact that about half of the study population had cardiac co-morbidity. This is a distinct advantage compared with estramustine phosphate and mitoxantrone in elderly patients, who may require prolonged treatment.

New compounds like epothilone B analogues are under evaluation in this setting [27Go], and also in patients who progress after taxane-based treatment; however, this new class of non-taxane tubulin polymerising agents shows a high rate of neurotoxicity. Therefore, VRL definitely has its place in the therapeutic arsenal of HRPC.

In conclusion, intravenous VRL in combination with low-dose hydrocortisone represents a valid alternative therapeutic option for the treatment of patients with HRPC, especially those who are not suitable for treatment with taxanes and/or mitoxantrone, or after taxane failure. Its mild toxicity profile makes this regimen easily manageable in elderly patients, including those with cardiac co-morbidity. Patient convenience could be further improved by the use of oral VRL, which has been shown to be a helpful alternative to the intravenous form in patients with NSCLC [28Go] and advanced breast cancer [29Go]. The recognition of active and tolerable chemotherapy regimens will permit their sequential use, as patients with HRPC may have relatively prolonged survival. This recognition may also result in the development of combination chemotherapy regimens, and their use with biological therapies.


    Acknowledgements
 
This study was supported by Institut de Recherche Pierre Fabre, Boulogne, France.

Received for publication March 12, 2004. Revision received June 21, 2004. Accepted for publication June 22, 2004.


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Scher H, Steineck G, Kelly WK. Hormone-refractory prostate cancer: refining the concept. Urology 1995; 46: 142–148.[CrossRef][ISI][Medline]

2. Rago R. Management of hormone-sensitive and hormone refractory metastatic prostate cancer. Cancer Control 1998; 5: 513–521.[Medline]

3. Tannock I, Osoba D, Stockler M et al. Chemotherapy with mitoxantrone plus prednisone or prednisone alone for symptomatic hormone-resistant prostate cancer: a Canadian randomized trial with palliative endpoints. J Clin Oncol 1996; 14: 1756–1764.[Abstract]

4. Culine S, Droz JP. Chemotherapy in advanced androgen-independent prostate cancer 1990-1999: a decade of progress? Ann Oncol 2000; 11: 1523–1530.[Abstract]

5. Fields-Jones S, Koletsky A, Wilding G et al. Improvements in clinical benefit with vinorelbine in the treatment of hormone-refractory prostate cancer: a phase II trial. Ann Oncol 1999; 10: 1307–1310.[Abstract]

6. Oudard S, Caty A, Humblet Y et al. Phase II study of vinorelbine in patients with androgen-independent prostate cancer. Ann Oncol 2001; 12: 847–852.[Abstract]

7. Morant RH, Schmitz SF, Bernhard J et al. Vinorelbine in androgen-independent metastatic prostate carcinoma–a phase II study. Eur J Cancer 2002; 38: 1626–1632.[CrossRef][ISI][Medline]

8. The Elderly Lung Cancer Vinorelbine Italian Study Group. Effects of vinorelbine on quality of life and survival of elderly patients with advanced non-small-cell lung cancer. J Natl Cancer Inst 1999; 91: 66–72.[Abstract/Free Full Text]

9. Vogel C, O'Rourke M, Winer E et al. Vinorelbine as first-line chemotherapy for advanced breast cancer in women 60 years of age or older. Ann Oncol 1999; 10: 397–402.[Abstract]

10. Buccheri G, Ferrigno D. Vinorelbine in elderly patients with inoperable non-small cell lung carcinoma. Cancer 2000; 88: 2677–2685.[CrossRef][ISI][Medline]

11. Bubley G, Carducci M, Dahut W et al. Eligibility and response guidelines for Phase II clinical trials in androgen-independent prostate cancer: recommendations from the prostate-specific antigen working group. J Clin Oncol 1999; 17: 2506–2513.[Abstract/Free Full Text]

12. Burris HA 3rd, Moore MJ, Andersen J et al. Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer: a randomized trial. J Clin Oncol 1997; 15: 2403–2413.[Abstract]

13. World Health Organisation. WHO Handbook for Reporting Results of Cancer Treatment. Geneva: World Health Organisation 1979.

14. Scher H, Kelly K, Zhang ZF et al. Post-therapy serum prostate specific antigen level and survival in patients with androgen-independent prostate cancer. J Nal Cancer Inst 1999; 91: 244–251.[CrossRef]

15. Dowling AJ, Czaykowski PM, Krahn MD et al. Prostate specific antigen response to mitoxantrone and prednisone in patients with refractory prostate cancer: prognostic factors and generalizability of a multicenter trial to clinical practice. J Urol 2000; 163: 1481–1485.[ISI][Medline]

16. Laudis J, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: 159–174.[ISI][Medline]

17. Kantoff P, Halabi S, Conaway M et al. Hydrocortisone with or without mitoxantrone in men with hormone-refractory prostate cancer: results of the Cancer and Leukemia Group B 9182 study. J Clin Oncol 1999; 17: 3461–3467.[Abstract/Free Full Text]

18. Small E, Meyer M, Marshall E et al. Suramin therapy for patients with symptomatic hormone-refractory prostate cancer: results of a randomized phase III trial comparing suramin plus hydrocortisone to placebo plus hydrocortisone. J Clin Oncol 2000; 18: 1440–1450.[Abstract/Free Full Text]

19. Havlin K, Trump D. Aminoglutethimide: theoretical considerations and clinical results in advanced prostate cancer. Cancer Treat Res 1988; 39: 83–96.[Medline]

20. Kruit W, Stoter G, Klijn J et al. The efficacy of endocrine combination therapy with aminoglutethimide and hydrocortisone in metastatic prostatic cancer refractory to standard endocrine therapies. Eur J Cancer 2001; 37: (Abstr 804).

21. Petrylak DP, Macarthur RB, O'Connor J et al. Phase I trial of docetaxel with estramustine in androgen-independent prostate cancer. J Clin Oncol 1999; 17: 958–967.[Abstract/Free Full Text]

22. Oudard S, Banu E, Vannetzel JM et al. Results of a phase II randomized trial of docetaxel (D), estramustine (E) and prednisone (P)—two schedules—versus mitoxantrone (M) and prednisone (P) in patients with hormone-refractory prostate cancer (HRPC). Ann Oncol 2002; 13 (Suppl 5): 90 (Abstr 3250).[Free Full Text]

23. Sinibaldi V, Carducci MA, Moore-Cooper S et al. Phase II evaluation of docetaxel plus one-day oral estramustine phosphate in the treatment of patients with androgen-independent prostate carcinoma. Cancer 2002; 94: 1457–1465.[CrossRef][ISI][Medline]

24. Savarese DM, Halabi S, Hars V et al. Phase II study of docetaxel, estramustine and low-dose hydrocortisone in men with hormone-refractory prostate cancer: a final report of CALGB 9780. Cancer and Leukemia Group B. J Clin Oncol 2001; 19: 2509–2516.[Abstract/Free Full Text]

25. Petrylak DP, Tangen C, Hussain M et al. Randomized phase III trial of docetaxel (D)/estramustine (E) versus mitoxantrone (M)/prednisone (P) in men with androgen-independent prostate cancer (AIPCA). Proc Am Soc Clin Oncol 2004; 23: 2 (Abstr 2).

26. Eisenberger MA, De Wit R, Berry W et al. A multicenter phase III comparison of docetaxel (D)+prednisone (P) and mitoxantrone (MTZ)+P in patients with hormone-refractory prostate cancer (HRPC). Proc Am Soc Clin Oncol 2004; 23: 2 (Abstr 3).

27. Goodin S, Kane MP, Rubin EH et al. Epothilones: mechanism of action and biologic activity. J Clin Oncol 2004; 22: 2015–2025.[Abstract/Free Full Text]

28. Jassem J, Ramlau E, Karnicka-Modkowska H et al. A multicenter randomised phase II study of oral versus intravenous vinorelbine in advanced non-small cell lung cancer. Ann Oncol 2001; 12: 1375–1381.[Abstract]

29. Freyer G, Delozier T, Lichinister M et al. Phase II study of oral vinorelbine in first line advanced breast cancer chemotherapy. J Clin Oncol 2003; 21: 35–40.[Abstract/Free Full Text]