RESEA R C H Open Access
A randomized phase II trial of mitoxantrone,
estramustine and vinorelbine or bcl-2 modulation
with 13-cis retinoic acid, interferon and paclitaxel
in patients with metastatic castrate-resistant
prostate cancer: ECOG 3899
Robert S DiPaola
1*
, Yu-Hui Chen
2
, Mark Stein
1
, David Vaughn
3
, Linda Patrick-Miller
1
, Michael Carducci
4
,
Bruce Roth
5
, Eileen White
6
, George Wilding
7
Abstract
Background: To test the hypothesis that modulation of Bcl-2 with 13-cis retinoic acid (CRA)/interferon-alpha2b
(IFN) with paclitaxel (TAX), or mitoxantrone, estramustine and vinorelbine (MEV) will have clinical activity in men
with metastatic castrate-resistant prostate cancer (CRPC).
Methods: 70 patients were treated with either MEV (Arm A) in a 3-week cycle or CRA/IFN/TAX with an 8-week
cycle (Arm B). Patients were assessed for response, toxicity, quality of life (QOL), and the effect of treatment on Bcl-

noic acid (CRA) and interferon (IFN) enhanced the effect
* Correspondence: [email protected]
1
Department of Medicine, The Cancer Institute of New Jersey, UMDNJ-
RWJMS, New Brunswick NJ, USA
DiPaola et al. Journal of Translational Medicine 2010, 8:20
http://www.translational-medicine.com/content/8/1/20
© 2010 DiPaola et al; licensee BioMed Central Ltd. This is an O pen Access article distributed under the terms of the Creative Commons
Attribution Lic ense (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reprodu ction in
any medium, provided the original work is properly ci ted.
of paclitaxel chemotherapy, and that the combination can
be safely administered in phase I studies [5-7]. More
recently, our group and other investigators have studied
novel BH3 domain mimetics such as AT101 and
ABT263, which modulate multiple Bcl-2 family proteins,
and are being tested in early clinical studies [8-10].
The development of novel chemotherapeutic combina-
tions to overcome tumor resistance may also be impor-
tant [11,12]. In this regard, previous data demonstrated
activity with estramustine-based combinations and the
activity of combined navelbine and mitoxantrone, sup-
porting the testing of combination therapy with estra-
mustine, mitoxantrone, and vinorelbine in patients with
CRPC. One study demonstrated a potential clinical ben-
efit to adding estramustine with vinorelbine as second
line therapy in CRPC [13]. Another study conducted by
the Hellenic Cooperative Oncology Group, evaluated the
safety and activity of the combination in CRPC and
found evidence of activity including complete response s
in measurable disease [14].

metastases were required to h ave a PSA level of ≥ 20
ng/ml. Patients with soft tissue metastases and/or visc-
eral disease were required to have either measurable
disease or a PSA level of ≥ 20 ng/ml. Patients were
required to have prior treatment with bilateral orchiect-
omy or other primary hormonal therapy with evidence
of treatment failure (patients who had not undergone
bilateral orchiectomy were required to continue LHRH
agonist therapy while receiving protocol therapy). Fluta-
mide or Nilutamide must have been discontinued at
least 4 weeks prior to randomization, and bicalutamide
at least 6 weeks prior to randomization, with evidence
of progressive disease. Patients could not have had prior
chemotherapy or prior Strontium 89, Samarium 153, or
other radioisotope therapies. Patients must have had
adequate bone marrow function, defined as WBC ≥
4000/mm
3
, granulocytes ≥ 2000/mm
3
, platelet count ≥
100,000/mm
3
, bilirubin ≤ 1.5 mg/dl, SGOT (AST) and
SGPT (ALT) ≤ 2 times the institutional upper limit of
normal, creatinine ≤ 2.0 mg/ dl or a calculated creatinin e
clearance ≥ 50 ml/min, LVEF ≥ 50% a s proven by
MUGA within 4 weeks of study entry, no active angina
pectoris, or known heart disease of New York Heart
Association Class III-IV. Patients must have had no his-

subcutaneously, each
administered on days 1 and 2 of each week, followed by
paclitaxel 75 mg/m
2
intravenously on day 2 of each
week with an 8-week cycle (Arm B). Each 8-week course
consisted of 6 weeks of treatment followed by a 2-week
rest. Treatment was continued until evidence of disease
progression or unacceptable toxici ty. As prophylaxis for
hypersensitivity reactions, patients were premedicated
with dexamethasone 20 mg, cimetidine (or ranitidine)
300 mg (50 mg), and benadryl 50 mg intravenously 30
minutes prior to each dose of paclitaxel. All dosing was
DiPaola et al. Journal of Translational Medicine 2010, 8:20
http://www.translational-medicine.com/content/8/1/20
Page 2 of 9
based on patients’ actual weight at the beginning of each
cycle. Conditions for dose modification in the case of
toxicity were specified in the protocol.
Efficacy and Safety
Response and progression were assessed according to
RECIST and were determined by investigator assessment
of radiographs and PSA. All toxicities were graded
according to the Common Toxicity Criteria (CTC) ver-
sion 2.0. Three self-reported measures of QOL were
completed by patients in both arms: the Functional
Ass essment of Cancer Therapy, Prostate Cancer Versi on
4 (FACT-P), the Brief Pain Inventory- Short Form (BPI),
and the Schwartz Cancer Fatigue Scale (SCFS). FACT-P
consists of 27 core items, which a ssess patient function

Statistical Design and Data Analysis
Patients were equally randomized to either of the two
treatment regimens and stratified by extent of disease
(measurable disease vs. non-measurable disease and ele-
vated PSA). The primary endpoint was the proportion
of patients responding by 6 months. PSA response was
defined as PSA decline from baseline value by ≥ 50%, or
normalization of PSA (<0.2 ng/mL), confirmed by a sec-
ond measurement at least 1 week later. An underlying
true response rate of 40% was considered evidence that
the treatment merited further study. On the other hand,
an underlying true response rate of 10% would be of no
clinical interest.
A two-stage accrual plan was employed for this study.
First, 7 eligible p atients per arm were to be registered to
the study and assessed for PSA response at 6 months.
Accrual would continue while these patients were fol-
lowed for the primary endpoint. Upon assessment , if 1 or
more of the 7 eligible patients responded, accrual would
continue to 30 eligible patients (35 total) per arm. If none
of the first 7 eligible patients demonstrated response, the
study would halt and the other patients accrued by this
time would be assessed. A stochastic curtailment a lgo-
rithm was used to guide decision-making. For each
potential number of patients registered at the time
response assessment among the first 7 eligible patients
was complete, the number of stage 1 responses required
to reopen accrual (out of all patients enrolled at that
time) was identified in the protocol, based on an ad hoc
rule, along with the corresponding probability of stop-

who are alive without progression). For patients with
DiPaola et al. Journal of Translational Medicine 2010, 8:20
http://www.translational-medicine.com/content/8/1/20
Page 3 of 9
measurable disease who did not progress, time to pro-
gression was censored at the last measurable disease
assessment or bone scan, whichever came first. For
patients without measurable disease who did not pro-
gress, time to progression was censored at the last bone
scan or last PSA measurement, whichever came first.
Descriptive statistics were used to characterize patients
at study entry. Patient demographics, adverse events,
and response rates were compared using Fisher’s exact
test. The method of Kaplan and Meier was used to char-
acterize overall survival and progression-free survival.
The stratified log rank test was used to test for differ-
ences in ove rall survival and progression-free survival by
treatment. The Wilcoxon signed-rank test was used to
test the differences in Bcl-2 levels between Day 1 Cycle
1 and Day 3 Cycle 1. All p-values are two sided.
In order to evaluate the effects of treatment on QOL,
changes in the FACT-P, TOI, BPI, and SCFS from base-
line to week 9/10 were calculated for each patient, and
evaluated using the Wilcoxon signed-rank test . Three
patients with measurable disease were stratified incor-
rectly to the non-measurable disease and elevated PSA
stratum (2 Arm A patients and 1 Arm B pati ent), while
1 Arm A patient without measurable disease was strati-
fied incorrectly to the measurable dis ease stratum. All
patients were analyzed according to their actual extent

Metastatic Disease
Lung 4 13% 4 13% 8 13%
Liver 3 9% 3 10% 6 10%
Bone 28 88% 29 94% 57 90%
Bone Marrow 0 0% 1 3% 1 2%
Pleura 1 3% 0 0% 1 2%
Other 11 34% 9 29% 20 32%
Prior Treatment
Orchiectomy 10 31% 7 23% 17 27%
Prostatectomy 17 53% 15 48% 32 51%
Other Surgery 19 59% 20 65% 39 62%
Radiation Therapy 18 56% 20 65% 38 60%
Hormonal Therapy 31 97% 29 94% 60 95%
Biologic Response Modifier 4 13% 2 6% 6 10%
DiPaola et al. Journal of Translational Medicine 2010, 8:20
http://www.translational-medicine.com/content/8/1/20
Page 4 of 9
with a range of 45 to 89 years. Most patients (95%) had
ECOG Performance Status of 0 or 1. About 90% of
patients were Caucasian. All patients had metastatic dis-
ease, most commonly in the bone (90%). Ninety-five
percent had been treated with hormonal therapy, and
27% had prior orchiect omy. As shown in Table 1, base-
line demographic and tumor characteristics were gener-
ally well balanced between the two arms, and no
significant difference between the two arms was found.
Adverse Events
Patients received a median of 5 and 2 cycles (15 and 16
weeks total) of protocol therapy on Arm A, which used
3-week cycles, and Arm B, which used 8-week cycles,

measurable disease were 14% on Arm A and 15% on
Arm B.
At the time of analysis, 60 patients have died and 3
patients are s till alive a mong eligible patients. Median
follow- up among patients still alive i s 59 month s. Figure
1 shows overall survival by treatment. Median survival is
19.4monthsforArmAand13.9monthsforArmB.
There was no statistically significant difference in overall
survival by treatment while controlling for extent of
disease (p = 0.42), but the study was not powered to
detect such difference. Figure 2 shows progression-free
survival by treatment. Median progression-free survival
is 5.9 months for Arm A and 2.5 months for Arm B.
There was no statistically significant difference in pro-
gression-free survival acro ss treatments aft er adjusting
for extent of disease (p = 0.34), but the study was not
powered to detect such difference.
Bcl-2 Levels
Among the 63 eligible patients, 18 and 16 patients had
Bcl-2 levels available on Arm A and Arm B, respectively.
For Arm A, the average Bcl-2/actin ratio was 0.88 and
1.32 on day 1 and day 3 of cycle 1, respectively, and the
difference was not significant (p = 0.47). For Arm B, the
average Bcl-2/actin ratio was 1.55 an d 1.00 on day 1
and day 3 of cycle 1, respectively, and the difference was
statistically significant (p = 0.03). Although the number
ass essed was small, no significant association was found
between Bcl-2 response and PSA response on either
arm.
Quality of Life

DiPaola et al. Journal of Translational Medicine 2010, 8:20
http://www.translational-medicine.com/content/8/1/20
Page 5 of 9
Table 2 Treatment-Related Toxicities (Patient Number)
Arm A (n = 35) Arm B (n = 35) Arm A (n = 35) Arm B (n = 35)
Grade Grade Grade Grade
Toxicity Type 1, 2 3 4 1, 2 3 4 Toxicity Type 1, 2 3 4 1, 2 3 4
Allergic reaction - - - - 2 - Stomatitis 5 - - 6 1 -
Allergic rhinitis 1 - - 1 - - Taste disturbance 6 - - 4 - -
Hemoglobin 31 3 1 28 5 - Vomiting 10 - - 9 1 -
Leukocytes 9 18 8 22 4 2 Diarrhea 5 2 - 11 - -
Neutrophils 4 11 18 16 7 2 GI-other 2 - - - - -
Platelets 20 5 - 13 1 - Hematuria - 1 - 1 - -
Transfusion: - 1 - - 4 - Rectal bleeding - - - - 1 -
Sinus tachycardia - 1 - - 1 - Alkaline phos 13 2 - 15 3 -
Supraventricular - - - - 1 - Bilirubin 1 1 1 5 - -
Cardiac-ischemia - - - 1 - - Hypoalbuminemia - - - 1 - -
Cardiac-left vent - 1 - - - - SGOT 7 - - 10 1 -
Edema 9 - - 4 - - SGPT 6 - - 7 1 -
Hypertension 2 - - 1 - - Febrile neutropenia - 4 - - 1 -
Hypotension 1 - - 3 - - Infection w/neutro - 1 1 - - -
Phlebitis 1 - - - - - Infectionw/oneutro 2 - 1 1 -
Thrombosis/emb - 6 1 - 1 - Hypercalcemia 1 - - 1 - -
Fatigue 21 7 1 23 7 1 Hyperglycemia 2 1 - 3 1 -
Fever - - - 7 1 - Hyperkalemia 1 - - 1 - -
Rigors/chills 1 - - 10 - - Hypertrig 6 - - 22 1 -
Sweating - - - 2 - - Hyponatremia 1 - - 1 - -
Weight gain 1 - - - - - Metabolic-other 1 - - - - -
Weight loss 14 - - 7 2 - Muscle weakness 1 - - - 1 -
PTT - 1 - - - - Ataxia - - - 1 - -

http://www.translational-medicine.com/content/8/1/20
Page 7 of 9
studies demonstrating an effect of CRA/IFN on Bcl-2
regulation [3-6]. We found a modest PSA response rate
of 23% and a progression free survival of only 2.5
months. Potential reasons for such modest activity m ay
include the lack of effect by CRA/IFN, the choice of tax-
ane, or limited therapy due to excessive toxicity. As
shown in Figure 3, patients had significantly lower QOL
scores, consistent with a clinica lly meaningful decline in
QOL at week 9/10 c ompared with baseline [16]. How-
ever, Bcl-2/actin ratios decreased in PBMCs with ther-
apy, supporting further study of this as a marker of drug
effect with Bcl-2 modulating therapi es. In fact, our cen-
ter and other investigators are now developing novel
agents such as the BH3 domain mimetics, which modu-
late multiple Bcl-2 family proteins [8-10].
The treatment arm with vinorelbine, mitoxantrone
and estramustine was designed based on data demon-
strating activity of this combination in limited studies.
We found a PSA response rate of 50%, progression free
survival of 5.9 months, and overall survival of 19.4
months, consistent with clinical activity of thi s regimen.
This is in agreement with a prior study of this regimen,
Figure 3 Changes in quality of life from baseline to week 9/10 (day 2 of cycle 4 for Arm A and day 1 of cycle 2 for Arm B).
DiPaola et al. Journal of Translational Medicine 2010, 8:20
http://www.translational-medicine.com/content/8/1/20
Page 8 of 9
which demonstrated a 56% PSA response, median dura-
tion of response of 6.9 months, and survival of 14.5

Vanderbilt University, Nashville, Tennessee, USA.
6
Department of Molecular
Biology and Biochemistry, Rutgers University, Piscataway, NJ, USA.
7
Department of Medicine, University of Wisconsin, Madison, WI, USA.
Authors’ contributions
RSD, YC, MS, MC, BR, EW, and GW contributed to enrollment, conception,
design, interpretation of data and reading and approval of the final
manuscript. LPM completed analysis and interpretation of quality of life data
and approved of the final manuscript. DV contributed to enrollment,
interpretation of data and reading and approval of the final manuscript. YC
also performed all statistical analysis.
Competing interests
The authors declare that they have no competing interests.
Received: 1 November 2009
Accepted: 24 February 2010 Published: 24 February 2010
References
1. Jemal A, Siegel R, Ward E, Hao Y, Xu J, Murray T, Thun MJ: Cancer statistics,
2008. CA Cancer J Clin 2008, 58:71-96.
2. Petrylak DP, Tangen CM, Hussain MH, Lara PN Jr, Jones JA, Taplin ME,
Burch PA, Berry D, Moinpour C, Kohli M, et al: Docetaxel and estramustine
compared with mitoxantrone and prednisone for advanced refractory
prostate cancer. N Engl J Med 2004, 351:1513-1520.
3. McDonnell TJ, Troncoso P, Brisbay SM, Logothetis C, Chung LW, Hsieh JT,
Tu SM, Campbell ML: Expression of the protooncogene bcl-2 in the
prostate and its association with emergence of androgen-independent
prostate cancer. Cancer Res 1992, 52:6940-6944.
4. Banerjee PP, Banerjee S, Brown TR: Bcl-2 protein expression correlates
with cell survival and androgen independence in rat prostatic lobes.

13. Nakabayashi M, Ling J, Xie W, Regan MM, Oh WK: Response to vinorelbine
with or without estramustine as second-line chemotherapy in patients
with hormone-refractory prostate cancer. Cancer J 2007, 13
:125-129.
14. Samelis GF, Kalofonos H, Adamou A, Kosmides P, Skarlos D, Aravantinos G,
Kiamouris C, Adimchi O, Fountzilas G, Dimopoulos AM: The combination of
estramustine, vinorelbine, and mitoxantrone in hormone-refractory
prostate cancer: a Phase II feasibility study conducted by the Hellenic
Cooperative Oncology Group. Urology 2005, 66:382-385.
15. Patrick-Miller LJ: Is there a role for the assessment of health-related
quality of life in the clinical evaluation of novel cytostatic agents?:
Commentary re: P. M. LoRusso, Improvements in quality of life and
disease-related symptoms in phase I trials of the selective oral
epidermal growth factor receptor tyrosine kinase inhibitor ZD1839 in
non-small cell lung cancer and other solid tumors. Clin. Cancer Res., 9:
2040-2048, 2003. Clin Cancer Res 2003, 9:1990-1994.
16. Jaeschke R, Singer J, Guyatt GH: Measurement of health status.
Ascertaining the minimal clinically important difference. Control Clin Trials
1989, 10:407-415.
doi:10.1186/1479-5876-8-20
Cite this article as: DiPaola et al.: A randomized phase II trial of
mitoxantrone, estramustine and vinorelbine or bcl-2 modulation with
13-cis retinoic acid, interferon and paclitaxel in patients with metastatic
castrate-resistant prostate cancer: ECOG 3899. Journal of Translational
Medicine 2010 8:20.
Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges