Articles

Abiraterone plus prednisone added to androgen deprivation

therapy and docetaxel in de novo metastatic castration-
sensitive prostate cancer (PEACE-1): a multicentre, open-
label, randomised, phase 3 study with a 2 × 2 factorial design
Karim Fizazi, Stéphanie Foulon, Joan Carles, Guilhem Roubaud, Ray McDermott, Aude Fléchon, Bertrand Tombal, Stéphane Supiot,
Dominik Berthold, Philippe Ronchin, Gabriel Kacso, Gwenaëlle Gravis, Fabio Calabro, Jean-François Berdah, Ali Hasbini, Marlon Silva,
Antoine Thiery-Vuillemin, Igor Latorzeff, Loïc Mourey, Brigitte Laguerre, Sophie Abadie-Lacourtoisie, Etienne Martin, Claude El Kouri,
Anne Escande, Alvar Rosello, Nicolas Magne, Friederike Schlurmann, Frank Priou, Marie-Eve Chand-Fouche, Salvador Villà Freixa,
Muhammad Jamaluddin, Isabelle Rieger, Alberto Bossi, on behalf of the PEACE-1 investigators*

Summary
Background Current standard of care for metastatic castration-sensitive prostate cancer supplements androgen Lancet 2022; 399: 1695–07
deprivation therapy with either docetaxel, second-generation hormonal therapy, or radiotherapy. We aimed to Published Online
evaluate the efficacy and safety of abiraterone plus prednisone, with or without radiotherapy, in addition to April 8, 2022
https://doi.org/10.1016/
standard of care.
S0140-6736(22)00367-1
See Comment page 1670
Methods We conducted an open-label, randomised, phase 3 study with a 2 × 2 factorial design (PEACE-1) at
*Members are listed in the
77 hospitals across Belgium, France, Ireland, Italy, Romania, Spain, and Switzerland. Eligible patients were male, appendix (p 2)
aged 18 years or older, with histologically confirmed or cytologically confirmed de novo metastatic prostate Department of Cancer
adenocarcinoma, and an Eastern Cooperative Oncology Group performance status of 0–1 (or 2 due to bone pain). Medicine (Prof K Fizazi MD),
Participants were randomly assigned (1:1:1:1) to standard of care (androgen deprivation therapy alone or with Department of Biostatistics
intravenous docetaxel 75 mg/m² once every 3 weeks), standard of care plus radiotherapy, standard of care plus and Epidemiology
(S Foulon MD), Oncostat
abiraterone (oral 1000 mg abiraterone once daily plus oral 5 mg prednisone twice daily), or standard of care plus U1018, Inserm, Ligue Contre le
radiotherapy plus abiraterone. Neither the investigators nor the patients were masked to treatment allocation. The Cancer (S Foulon), and
coprimary endpoints were radiographic progression-free survival and overall survival. Abiraterone efficacy was Department of Radiotherapy
first assessed in the overall population and then in the population who received androgen deprivation therapy (A Bossi MD), Institut Gustave
Roussy, University of Paris-
with docetaxel as standard of care (population of interest). This study is ongoing and is registered with Saclay, Villejuif, France; Vall
ClinicalTrials.gov, NCT01957436. d’Hebron Institute of
Oncology, Vall d’Hebron
Findings Between Nov 27, 2013, and Dec 20, 2018, 1173 patients were enrolled (one patient subsequently withdrew University Hospital,
Barcelona, Spain
consent for analysis of his data) and assigned to receive standard of care (n=296), standard of care plus radiotherapy (Prof J Carles MD); Institut
(n=293), standard of care plus abiraterone (n=292), or standard of care plus radiotherapy plus abiraterone (n=291). Bergonié 229 cours de
Median follow-up was 3·5 years (IQR 2·8–4·6) for radiographic progression-free survival and 4·4 years (3·5–5·4) l’Argonne, Bordeaux, France
(G Roubaud MD); Cancer Trials
for overall survival. Adjusted Cox regression modelling revealed no interaction between abiraterone and
Ireland, St Vincent’s University
radiotherapy, enabling the pooled analysis of abiraterone efficacy. In the overall population, patients assigned to Hospital, Dublin, Ireland
receive abiraterone (n=583) had longer radiographic progression-free survival (hazard ratio [HR] 0·54, 99·9% CI (Prof R McDermott MD); Centre
0·41–0·71; p<0·0001) and overall survival (0·82, 95·1% CI 0·69–0·98; p=0·030) than patients who did not receive Léon Bérard, Lyon, France
(A Fléchon MD); Clinique
abiraterone (n=589). In the androgen deprivation therapy with docetaxel population (n=355 in both with
Universitaire St Luc, Brussels,
abiraterone and without abiraterone groups), the HRs were consistent (radiographic progression-free survival 0·50, Belgium (Prof B Tombal MD);
99·9% CI 0·34–0·71; p<0·0001; overall survival 0·75, 95·1% CI 0·59–0·95; p=0·017). In the androgen deprivation Institut de Cancérologie de
therapy with docetaxel population, grade 3 or worse adverse events occurred in 217 (63%) of 347 patients who l’Ouest, René Gauducheau,
Saint-Herblain, France
received abiraterone and 181 (52%) of 350 who did not; hypertension had the largest difference in occurrence
(Prof S Supiot MD); Centre
(76 [22%] patients and 45 [13%], respectively). Addition of abiraterone to androgen deprivation therapy plus Pluridisciplinaire d’Oncologie,
docetaxel did not increase the rates of neutropenia, febrile neutropenia, fatigue, or neuropathy compared with Centre Hospitalier Universitaire
androgen deprivation therapy plus docetaxel alone. Vaudois, Lausanne, Switzerland
(D Berthold MD); Centre
Azureen de Cancerologie,
Interpretation Combining androgen deprivation therapy, docetaxel, and abiraterone in de novo metastatic castration- Mougins, France
sensitive prostate cancer improved overall survival and radiographic progression-free survival with a modest increase (P Ronchin MD); Amethyst
in toxicity, mostly hypertension. This triplet therapy could become a standard of care for these patients. Radiotherapy Center, Iuliu
Hatieganu University of
Medicine and Pharmacy,
Funding Janssen-Cilag, Ipsen, Sanofi, and the French Government. Cluj-Napoca, Romania
(G Kacso MD); Institut
Copyright © 2022 Elsevier Ltd. All rights reserved.

www.thelancet.com Vol 399 April 30, 2022 1695

 Articles

Paoli-Calmettes, Aix-Marseille
Université, CRCM,
Research in context
Marseille, France (G Gravis MD);
Evidence before this study Added value of this study
San Camillo Forlanini Hospitals,
Rome, Italy (F Calabro MD); We searched PubMed from Jan 1, 1984, to Dec 31, 2012, for To our knowledge, PEACE-1 is the first trial to show that
Clinique Sainte Marguerite, papers in English, using the terms “prostate cancer”, a triple systemic therapy—consisting of ADT, docetaxel,
Hyeres, France (J-F Berdah MD); “metastases”, and “phase 3 trial”, and found 16 studies. When and a second-generation androgen signalling inhibitor
Clinique Pasteur,
the PEACE-1 trial began in 2013, androgen deprivation (abiraterone, combined with prednisone)—improves both
Brest, France(A Hasbini MD);
Centre François Baclesse, Caen, therapy (ADT) had for several decades been the standard of radiographic progression-free survival and overall survival in
France (M Silva MD); CHU Jean care for men with metastatic castration-sensitive prostate patients with de novo mCSPC, without excessively increasing
Minjoz, Besancon, France cancer (mCSPC). All randomised trials conducted before the toxicity (mostly hypertension and aminotransferase increase).
(A Thiery-Vuillemin MD);
mid-2010s had not shown improvement in outcomes with
Clinique Pasteur, Toulouse, Implications of all the available evidence
France (I Latorzeff MD); IUCT— new therapies compared with ADT. During recruitment for the
The findings from the PEACE-1 trial, combined with the evidence
Oncopole, Toulouse, France PEACE-1 trial (2013–18), overall survival of patients with de
(L Mourey MD); Centre Eugène from other studies, add weight to the concept that systemic
novo mCSPC was shown to be improved by combining ADT
Marquis, Rennes, France treatment intensification provides clinical benefit for men with
with either docetaxel or one of the second-generation
(B Laguerre MD); Institut de mCSPC and illustrate that early treatment intensification is more
Cancerologie de l’Ouest, androgen receptor axis inhibitors (abiraterone, apalutamide,
effective than identical or similar treatments being used
Angers, France or enzalutamide) and radiotherapy to the primary tumour, in
(S Abadie-Lacourtoisie MD); sequentially when the disease has become castration-resistant.
men with low metastatic burden. It had not yet been shown
Centre Georges-François At a minimum, patients with mCSPC with a high metastatic
whether combining some of these new therapies with ADT in
Leclerc, Dijon, France burden who are fit enough to be treated with docetaxel should
(E Martin MD); Centre Catherine triple therapy could provide further clinical benefits.
be considered for this triple systemic therapy.
de Sienne, Nantes, France
(C El Kouri MD); Strasbourg
Oncologie Libérale, Strasbourg,
France (A Escande MD); Institut Introduction or cytologically confirmed prostate adenocarcinoma
Català d’Oncologia, Hospital After several decades with no substantial progress, the documented as de novo metastatic by bone scan, CT, or
Universitari Josep Trueta, treatment of de novo (synchronous) metastatic prostate MRI (according to Response Evaluation Criteria in Solid
Girona, Spain (A Rosello MD);
cancer has drastically evolved in the past 10 years Tumors [RECIST] version 1.1 criteria) and an Eastern
Institut de Cancérologie Lucien
Neuwirth, St Priest en Jarez, with the addition of diverse treatments to androgen Cooperative Oncology Group (ECOG) performance status
France (N Magne MD); CHIC deprivation therapy (ADT), the former standard of care of 0–1 (or 2 due to bone pain) were eligible for inclusion,
Quimper, Quimper, France (SOC). The prognosis for men with de novo metastatic provided they had received ADT for no more than
(F Schlurmann MD); CHD
castration-sensitive prostate cancer (mCSPC) has been 3 months before randomisation and had at least 6 weeks
Vendée, La Roche sur Yon,
France (F Priou MD); Centre improved by using ADT concomitantly with either between the initiation of ADT and the first docetaxel
Antoine Lacassagne, Nice, docetaxel1–5 or one of the second-generation androgen dose. Patients with a pure small-cell carcinoma, or
France (M-E Chand-Fouche MD); receptor axis inhibitors (abiraterone,6–8 apaluta­mide,9 or previous prostate cancer treated by a definitive local
Institut Català d’Oncologia,
Cap de Servei Oncologia
enzalutamide10,11). In 2018, radiotherapy to the primary treatment were not eligible. All participants provided
Radioteràpica, Hospital tumour was also shown to extend the overall survival of written informed consent before any study-specific
Universitari Germans Trias, patients with low-volume metastatic burden.12 The procedures or randomisation were initiated.
Badalona, Catalunya, Spain benefits of these various combinatory therapies were The initial protocol was reviewed and first approved on
(S V Freixa MD); Cork University
Hospital, Cork, Ireland
confirmed by meta-analyses13–15 and have helped to shape July 10, 2013, by the French Independent Ethics
(M Jamaluddin MD); Unicancer, the current guidelines for the treatment of mCSPC.16–18 Committee (Comité de Protection des Personnes Ile de
Paris, France (I Rieger MSc) Nevertheless, it remains to be established whether France VII). The protocol complied with the ethical
Correspondence to: merging some of these new therapies could provide principles of the Declaration of Helsinki and was
Prof Karim Fizazi, Department of further clinical benefits and, if so, what combination approved by Institutional Review Boards at each study
Cancer Medicine, Institut
might be most suitable for de novo mCSPC.19 Conducted site. The trial conformed with the International
Gustave Roussy, University of
Paris-Saclay, Villejuif 94805, by a European consortium (PEACE),20 this study aimed to Conference on Harmonization and Good Clinical
France evaluate the efficacy and safety of abiraterone plus Practice guidelines, and applicable regulatory require­
karim.fizazi@gustaveroussy.fr prednisone, with or without radiotherapy, in addition to ments. Steering and Independent Data Monitoring
See Online for appendix SOC (ADT with or without docetaxel). Committees roles and members are described in the
appendix (pp 3–4).
Methods Two major protocol amendments were made during
Study design and participants the accrual period, to account for the evolution of SOC
We conducted an open-label, randomised, active- for men with mCSPC. Although ADT alone was the SOC
controlled, phase 3 study with a 2 × 2 factorial design when the trial began recruitment in 2013, the protocol
(PEACE-1) at 77 sites across seven European countries was amended on Oct 5, 2015 (273 patients had already
(Belgium, France, Ireland, Italy, Romania, Spain, and been accrued), to allow for docetaxel use in SOC
Switzerland). The complete list of inclusion and exclusion worldwide, after ADT with docetaxel had shown
criteria is provided in the appendix (pp 5–6). Briefly, male improvement in overall survival.2,3 Then, in 2017, ADT
patients aged 18 years or older with histologically confirmed plus abiraterone was also shown to improve overall

1696 www.thelancet.com Vol 399 April 30, 2022

 Articles

survival compared with ADT alone.6,7 Thereby, an had to be administered within 14 days after randomisation
additional amendment was filed on Aug 10, 2017, to make and be at least 6 weeks after ADT initiation. Patients
docetaxel mandatory for the remainder of the patients to assigned to receive abiraterone received 1000 mg of
be accrued, so that the trial could evaluate the addition of abiraterone (four 250 mg tablets, orally) once daily plus
abiraterone to ADT with docetaxel (given that the prednisone 5 mg orally twice daily, starting within 6 weeks
evaluation of ADT plus abiraterone compared with ADT after ADT initiation. PEACE-1 began accrual in 2013,
alone had already been addressed by other trials). Further before the LATITUDE6 and the STAMPEDE7 abiraterone
details on the amendments to the protocol and statistical trials were reported and before abiraterone was approved
analysis plan are provided in the appendix (pp 7–8). for mCSPC, which is why we used 10 mg of prednisone,
Evolution of the sample size and allocation of patients similar to what is approved for metastatic castration-
according to these key amendments is depicted in the resistant prostate cancer (mCRPC). Abiraterone and
appendix (pp 9–10). prednisone (hereafter referred to as abiraterone) were
administered until disease progression to castration
Randomisation and masking resistance, withdrawal of consent, unacceptable toxicity,
Eligible patients were centrally randomly assigned in the or death. Patients assigned to receive radiotherapy
Alea Clinical Portal), in a 1:1:1:1 ratio, to SOC, SOC plus (74 Gy in 37 fractions administered over 7–8 weeks) For more on the Alea Clinical
radiotherapy, SOC plus abiraterone (abiraterone plus were planned to start radiotherapy at least 3 weeks Portal see https://www.
aleaclinical.eu
prednisone), or SOC plus radiotherapy plus abiraterone. after docetaxel completion (but no more than 8 weeks
This randomisation process was performed via the after completion). Treatments given after disease
Tenalea autonomous software, solely accessed by the trial progression followed common practice, and investigators
data manager within the Epidemioloy and Biostatistics were able to change therapy as they thought was most
facility of the Gustave Roussy Center in Villejuif, France appropriate, on the basis of prostate-specific antigen
until 2015, then by each investigator. (PSA) concen­ tration variation or clinical progression,
Randomisation was done using a minimisation even in absence of radiographical evidence. The treatment
algorithm, stratified by study site, ECOG performance flow chart is depicted in the appendix (p 10).
status (0 vs 1–2), type of ADT (gonadotropin-releasing Every enrolled patient was to be followed-up for a
hormone agonist vs antagonist vs bilateral orchiectomy), duration of 10 years. The complete list of assessments
planned administration of docetaxel (yes vs no), and performed at each visit and follow-up intervals are
disease extent or burden based on metastatic status detailed in the appendix (pp 12–13). Survival status data
(lymph node metastases only vs bone metastases [with or were gathered within 3 months of the cutoff date for
without lymph node metastases] vs visceral metastases). 97% of patients.
A similar weight was given to each of the five
stratification factors, with 80% probability to minimise Outcomes
imbalance in the number of patients assigned to each The coprimary endpoints were radiographic progression-
treatment group. As the case report form gathered free survival and overall survival. Radiographical pro­
information for patient stratification according to gression of soft-tissue lesions was evaluated by either CT
high-volume versus low-volume metastatic burden or MRI, on the basis of RECIST version 1.1. Progression
classification (as defined in the CHAARTED study2 [high of bone lesions was assessed by bone scan according to
volume was defined as the presence of visceral the adapted version of Prostate Cancer Working Group 2
metastases or at least four bone lesions with one beyond criteria,21 with no secondary bone scan required to
the vertebral bodies and pelvis]), this later classification confirm progression. Radiographic progression-free
was retained to enable inter-trial comparisons. No survival was defined as the time between randomisation
masking was performed in this study. and the occurrence of radiographical progression or
death from any cause. Overall survival was defined as the
Procedures time between randomisation and death from any cause.
In all patients in the study, ADT was planned to be Patients without events were censored at the date of last
continuously maintained by either a gonadotropin- follow-up.
releasing hormone agonist or antagonist, or bilateral The secondary endpoints were castration-resistant
orchiectomy. Patients assigned to receive docetaxel as part prostate cancer (CRPC)-free survival, serious-genito­
of SOC with continuous ADT were to receive six cycles of urinary-event-free survival, prostate-cancer-specific survival,
intravenous docetaxel (75 mg/m² per cycle; maximum time to next skeletal-related event, PSA response rate,
dose of 150 mg per cycle), to be administered every prognostic study of serum PSA measured 6–8 months
3 weeks (plus or minus 3 days). Granulocyte colony- after initiation of systemic therapy, time to pain pro­
stimulating factor (G-CSF) injections after each docetaxel gression, time to chemotherapy for CRPC, quality of life,
cycle were recommended until the protocol amendment changes in bone mineral density, correlation of biomarkers
on Jan 22, 2018, made G-CSF prophylaxis mandatory for with outcome, event rate per 100 person-years of treatment
patients who received docetaxel. The first docetaxel cycle analysis, and toxicity. CRPC-free survival was defined as

www.thelancet.com Vol 399 April 30, 2022 1697

 Articles

the time between randomisation and CRPC or death from This trial had a factorial design, based on the
any cause. CRPC was defined as either radiographical assumption that there would be no interaction between
progression or a confirmed PSA rise (based on three abiraterone and radiotherapy. Before analysing the
indepen­dent measurements: A, B, and C, with A<B<C, various outcomes, the presence of such an interaction
and C≥0·50 ng/mL), with a serum testosterone within for both coprimary endpoints was tested by analyses of
castrated range (<0·50 ng/mL). Prostate-cancer-specific maximum likelihood estimates using a Cox model
survival was defined as the time from randomisation to the adjusted for stratification factors (ECOG performance
occurrence of death from prostate cancer. The event rate status [0 vs 1–2], ADT type [gonadotropin-releasing
per 100 person-years of treatment was the number of hormone agonist vs antagonist vs bilateral orchiectomy],
adverse events divided by the amount of person-time metastatic burden [low vs high; as defined in the
observed (ie, 100 patients experiencing any adverse event Randomisation and masking section],2 and docetaxel
over a 1-year period of exposure, with an exposure time [no docetaxel before amendment vs no docetaxel after
from randomisation to CRPC or last follow-up). Adverse amendment vs with docetaxel after amendment]). In the
events were graded on the basis of the National Cancer absence of a qualitative interaction (p>0·05) between
Institute Common Terminology Criteria for Adverse abiraterone and radiotherapy, the groups were to be
Events version 4.0. Safety analysis was based on the highest combined two by two on the basis of abiraterone
grade adverse event recorded during the period spanning administration, regardless of radiotherapy, before se­
from treatment initiation to CRPC. Apart from CRPC-free quentially assessing abiraterone efficacy first in the
survival, prostate-cancer-specific survival, event rate per overall population and then in the ADT with docetaxel
100 person-years, and toxicity, secondary endpoint results population (assuming the significance level was reached
were still under investigation at the time this manuscript in first instance). As the preplanned number of pro­
was submitted and so are not reported here. gression-free survival and overall survival events had
not been reached in the study population at the time
Statistical analysis this paper was written, the efficacy of radiotherapy
The initial protocol in 2013 planned for a sample size of remains to be analysed (appendix p 16). The coprimary
916 patients. However, due to a change in the SOC and the endpoint results presented here correspond to the final
knowledge gained from the LATITUDE and STAMPEDE planned analysis; no interim analysis was conducted.
trial results published in late 2017, the planned sample All efficacy analyses were conducted in the intention-to-
size was increased to 1173 and the coprimary study aims treat population, defined as all patients who were
were revised (appendix pp 7–8). Because PEACE-1 was the randomly assigned to a treatment group. Safety analyses
only phase 3 trial to date that could determine abiraterone were conducted in the safety population, according to
efficacy in patients with mCSPC receiving ADT plus the treatment actually received by the patients (those
docetaxel as SOC, the timing of the final analyses was who did not receive any investigational treatment were
based on this population (population of interest). Sample not included in the safety analyses).
size calculation was performed with East software (Cytel; The median follow-up was estimated by the inverse
Cambridge, MA, USA), on the basis of the assumption Kaplan-Meier method. Time-to-event endpoints were
that no significant interaction would take place between estimated by the Kaplan-Meier method. The Cox
radiotherapy and abiraterone, to allow for a 2 × 2 factorial proportional hazards model adjusted for radiotherapy
analysis of abiraterone efficacy. For abiraterone efficacy and stratification factors provided significances and an
evaluation, the predetermined acceptable probability of a estimate of the abiraterone effect (p value and HR with
type I error was set at 0·05, divided between the two CIs adjusted to match adjustment made to significance
coprimary endpoints (0·049 for overall survival and 0·001 levels in the corresponding test [ie, 99·9% for
for radiographic progression-free survival). We radiographic progression-free survival, 95·1% for overall
hypothesised that adding abiraterone to ADT plus survival, and 95% for secondary endpoints]). The
docetaxel would improve overall survival by 30% over a assumption of proportional hazards was evaluated on
median of 53 months and progression-free survival by the basis of the weighted Schoenfeld residuals. The
40% over 30 months. Hence, with 355 patients assigned to heterogeneity of the effect of abiraterone in the
ADT with docetaxel plus abiraterone (with or without predefined stratification subgroups was evaluated using
radiotherapy) and 355 patients assigned to ADT with Cox proportional hazards models adjusted for radio­
docetaxel without abiraterone (with or without therapy and stratification factors and including an
radiotherapy), 249 deaths would give an 80% power to interaction term between abiraterone and the studied
detect an overall survival hazard ratio (HR) of 0·70 at a stratification variable. These analyses are illustrated
two-sided α level of 0·049. With the same group allocation, using forest plots. Notably, we analysed the efficacy of
262 radiographic progression events or deaths were abiraterone in the ADT with docetaxel population by
predicted to have an 80% power to detect a radiographic disease burden. Non-parametric CIs for the median
pro­gression-free survival HR of 0·60 at a two-sided α level survival differences between groups were analysed by
of 0·001. the bootkm function of the R Hmisc package

1698 www.thelancet.com Vol 399 April 30, 2022

 Articles

(5000 repeats). Statistical analyses were performed with survival outcome considered, the median follow-up
SAS (version 9.4) and R (version 4.0.2). periods did not differ between the groups who did or did
This study is ongoing and is registered with not receive abiraterone. For the radiographic progression-
ClinicalTrials.gov, NCT01957436. free survival analyses, the median follow-up periods
from randomisation to the cutoff date of Sept 1, 2020,
Role of the funding source were 3·53 years for the overall population (3·52 years for
The funders of the study had no role in study design, the SOC without abiraterone groups and 3·56 years for
data collection, data analysis, data interpretation, or the SOC plus abiraterone groups; log-rank p=0·86) and
writing of the report. 2·99 years for the ADT with docetaxel population
(3·00 years and 2·97 years; log-rank p=0·93). For the
Results overall survival analyses, the median follow-up periods
Between Nov 27, 2013, and Dec 20, 2018, 1173 patients from randomisation to the cutoff date of June 1, 2021,
were enrolled (one patient subsequently withdrew were 4·41 years for the overall population (4·44 years for
consent for analysis of his data) and assigned to receive the SOC without abiraterone groups and 4·39 years for
SOC (n=296), SOC plus radiot­herapy (n=293), SOC plus the SOC plus abiraterone groups; log-rank p=0·98) and
abiraterone (n=292), or SOC plus radiotherapy plus 3·81 years for the ADT with docetaxel population
abiraterone (n=291; figure 1). SOC was ADT alone in (3·75 years and 3·85 years; log-rank p=0·95).
462 patients and ADT with docetaxel in 710 patients In the overall population (n=1172), no interaction
(table 1). between abiraterone and radiotherapy was found
For radiographic progression-free survival and overall for radiographic progression-free survival (p=0·64),
survival, the median follow-up periods were longer in overall survival (p=0·86), CRPC-free survival (p=0·56),
the overall population (3·5 years [IQR 2·8–4·6] and or prostate-cancer-specific survival (p=0·54), after
4·4 years [3·5–5·4], respectively) than in the ADT with adjusting for the four stratification factors. Likewise,
docetaxel population (3·0 years [2·1–3·8] and 3·8 years in the ADT with docetaxel population (n=710), no
[2·9–4·5], respectively). Regardless of the population and interaction between abiraterone and radiotherapy was

1173 patients enrolled and randomly assigned

1 withdrew consent for analysis of

their data, data not included

296 assigned to SOC 293 assigned to SOC plus 292 assigned to SOC plus 291 assigned to SOC plus
radiotherapy abiraterone abiraterone plus radiotherapy

178 received docetaxel 177 received docetaxel 177 received docetaxel 178 received docetaxel
118 did not receive docetaxel 116 did not receive docetaxel 115 did not receive docetaxel 113 did not receive docetaxel

589 in SOC (with or without radiotherapy) without abiraterone 583 in SOC (with or without radiotherapy) plus abiraterone groups
groups 337 alive with data from the past 3 months
306 alive with data from the past 3 months 18 alive without data from the past 3 months (lost to
15 alive without data from the past 3 months follow-up or consent withdrawn)
(lost to follow-up or consent withdrawn) 228 deceased at data cutoff
268 deceased at data cutoff

589 included in the overall population for efficacy analysis 583 included in the overall population for efficacy analysis
355 included in the ADT with docetaxel population for efficacy 355 included in the ADT with docetaxel population for efficacy
analysis analysis

350 included in the ADT with docetaxel population for safety 347 included in the ADT with docetaxel population for safety
analysis analysis
237 included in the ADT without docetaxel population for safety 226 included in the ADT without docetaxel population for safety
analysis analysis

Figure 1: Trial profile

ADT=androgen deprivation therapy. SOC=standard of care.

www.thelancet.com Vol 399 April 30, 2022 1699

 Articles

found for radiographic progression-free survival (p=0·94), In the overall population, the addition of abiraterone to
overall survival (p=0·85), CRPC-free survival (p=0·75), or SOC (with or without docetaxel and with or without
prostate-cancer-specific survival (p=0·98), after adjusting radiotherapy) decreased the number of radiographic
for ECOG performance status, ADT type, and disease progression events or deaths from 371 to 252, improved
burden. Consequently, the comparative evaluation of the median progression-free survival from 2·22 years
abiraterone efficacy on survival outcomes was conducted (IQR 1·09–6·03) to 4·46 years (1·72–not reached), and
by pooling the groups two by two (SOC without reduced the relative risk of radiographic progression or
abiraterone [with or without radiotherapy] groups vs SOC death by 46% compared with patients who did not receive
plus abiraterone [with or without radiotherapy] groups). abiraterone (adjusted HR for radiographic progression-
free survival 0·54, 99·9% CI 0·41–0·71; p<0·0001;
Overall population (n=1172) ADT with docetaxel population figure 2A, table 2). The addition of abiraterone to SOC
(n=710)* (with or without docetaxel and with or without
SOC plus SOC without SOC plus SOC without radiotherapy) decreased the number of deaths from 268
abiraterone groups abiraterone groups abiraterone groups abiraterone groups to 228, improved the median overall survival from 4·72
(with or without (with or without (with or without (with or without
years (IQR 2·59–not reached) to 5·72 years (2·72–not
radiotherapy; radiotherapy; radiotherapy; radiotherapy;
n=583) n=589) n=355) n=355) reached), and reduced the risk of death from any cause by
18% (adjusted HR for overall survival 0·82, 95·1% CI
Assigned to receive 291 (50%) 293 (50%) 178 (50%) 177 (50%)
radiotherapy 0·69–0·98; p=0·030; figure 2C, table 2). In the overall
Country population, the effect of abiraterone on radio­ graphic
Belgium 29 (5%) 25 (4%) 16 (5%) 16 (5%) progression-free survival and overall survival showed a
France 458 (79%) 462 (78%) 278 (78%) 280 (79%) high consistency across most of the predefined
Ireland 30 (5%) 30 (5%) 17 (5%) 13 (4%) subgroups, except for the under-represented subgroup of
Italy 1 (<1%) 3 (1%) 0 0
patients who had bilateral orchiectomy and those who
Romania 4 (1%) 5 (1%) 0 0
did not receive docetaxel based on the investigator’s
Spain 55 (9%) 56 (10%) 38 (11%) 39 (11%)
decision (figures 3A, 3C). For overall survival only, the
Switzerland 6 (1%) 8 (1%) 6 (2%) 7 (2%)
effect of the addition of abiraterone was pronounced in
patients with high-volume metastatic burden (figure 3C).
Age, years
As SOC plus abiraterone showed a superior efficacy
Median 67 (61–72) 66 (59–72) 66 (60–70) 66 (59–70)
compared with SOC without abiraterone that met the
Range 37–94 43–87 37–85 44–84
required predefined significance level (type I error of
ECOG performance status
0·049 for overall survival and 0·001 for radiographic
0 412 (71%) 412 (70%) 250 (70%) 246 (69%)
progression-free survival) within the overall population,
1–2 171 (29%) 177 (30%) 105 (30%) 109 (31%)
abiraterone efficacy was further investigated within the
T stage
more restricted ADT with docetaxel population to assess
T1 23 (4%) 23 (4%) 10 (3%) 13 (4%)
the preplanned coprimary endpoints. In this population,
T2 109 (19%) 94 (16%) 64 (19%) 45 (13%)
compared with SOC (including docetaxel) without
T3 287 (51%) 310 (53%) 167 (49%) 189 (55%)
abiraterone, the addition of abiraterone decreased the
T4 98 (17%) 99 (17%) 68 (20%) 65 (19%) number of radiographic progression events or deaths
Tx 45 (8%) 54 (9%) 32 (9%) 35 (10%) from 211 to 139, improved the median radio­ graphic
Missing data 21 (4%) 9 (2%) 14 (4%) 8 (2%) progression-free survival from 2·03 years (IQR 1·09–not
N stage reached) to 4·46 years (1·90–not reached), and reduced
N1 307 (55%) 325 (57%) 198 (58%) 207 (60%) the relative risk of radiographic progression or death by
N0 186 (33%) 174 (30%) 99 (29%) 97 (28%) 50% (adjusted HR for radiographic progression-free
NX 69 (12%) 76 (13%) 43 (13%) 39 (11%) survival 0·50, 99·9% CI 0·34–0·71; p<0·0001; figure
Missing data 21 (4%) 14 (2%) 15 (4%) 12 (3%) 2B, table 2). Similarly, compared with SOC (including
Time from diagnosis, months docetaxel) without abiraterone, the addition of abira­
Median 2·3 (1·6–3·2) 2·3 (1·4–3·1) 2·2 (1·6–3·0) 2·2 (1·4–2·9) terone reduced the number of deaths from 151 to 121,
Missing data 10 (2%) 10 (2%) 6 (2%) 7 (2%) improved the median overall survival from 4·43 years
Metastatic localisation (IQR 2·47–not reached) to not reached, and reduced the
Bone† 472 (81%) 475 (81%) 287 (81%) 279 (79%) relative risk of death from any cause by 25% (adjusted
Lymph node only 47 (8%) 52 (9%) 27 (8%) 29 (8%) HR for overall survival 0·75, 95·1% CI 0·59–0·95;
Visceral‡ 64 (11%) 62 (11%) 41 (12%) 47 (13%) p=0·017; figure 2D, table 2).
Metastatic burden§ By analysing the ADT with docetaxel population
High burden 331 (57%) 336 (57%) 224 (63%) 232 (65%) according to the disease burden, we observed that
Low burden 252 (43%) 253 (43%) 131 (37%) 123 (35%) compared with SOC (including docetaxel) without
(Table 1 continues on next page) abiraterone, the addition of abiraterone decreased the
number of radiographic pro­gression events or deaths

1700 www.thelancet.com Vol 399 April 30, 2022

 Articles

from 55 to 41 and from 156 to 97, and reduced the Overall population (n=1172) ADT with docetaxel population
(n=710)*
relative risk of radio­graphic progression or death by
42% and 53%, in patients with low metastatic burden SOC plus SOC without SOC plus SOC without
abiraterone groups abiraterone groups abiraterone groups abiraterone groups
and patients with high metastatic burden, respectively
(with or without (with or without (with or without (with or without
(low-volume burden median not reached vs 2·7 years; radiotherapy; radiotherapy; radiotherapy; radiotherapy;
adjusted HR 0·58, 99·9% CI 0·29–1·15; p=0·0061; n=583) n=589) n=355) n=355)
high-volume burden median 4·1 years vs 1·6 years; (Continued from previous page)
adjusted HR 0·47, 99·9% CI 0·30–0·72; p<0·0001; Gleason score
appendix p 10, table 2). Median overall survival in ≤7 145 (25%) 133 (23%) 79 (23%) 71 (20%)
patients with high metastatic burden improved from 8–10 429 (75%) 441 (77%) 270 (77%) 276 (80%)
3·47 years with SOC without abiraterone to 5·14 years Missing data 9 (2%) 15 (3%) 6 (2%) 8 (2%)
when abiraterone was added, corresponding to a 28%
PSA at randomisation, ng/mL
reduction in relative risk of death from any cause
Median 14 (3–62) 11 (3–55) 14 (2–59) 12 (3–60)
(adjusted HR 0·72, 95·1% CI 0·55–0·95; p=0·019;
Missing data 2 (<1%) 4 (1%) 0 2 (<1%)
figure 2F, table 2). As the data were not mature, no
Medical history
conclusive response in overall survival could be shown
Hypertension 270 (47%); N=574 241 (43%); N=562 156 (44%); N=352 148 (43%); N=344
for patients with low metastatic burden (figures 2E, 3D,
Type 2 diabetes 62 (11%); N=566 80 (14%); N=556 33 (9%); N=351 56 (16%); N=344
table 2).
High cholesterol 229 (40%); N=568 229 (41%); N=556 136 (39%); N=351 130 (38%); N=343
In the ADT with docetaxel population, compared with
SOC without abiraterone, the addition of abiraterone Data are n (%) or median (IQR) unless otherwise stated. All numbers are rounded to the nearest integer. Ethnicity-
related information are not presented, as French laws forbid the collection of such data. SOC in the overall population
delayed castration resistance (median CRPC-free survival was ADT with or without docetaxel. SOC in the ADT with docetaxel population was ADT with docetaxel.
1·45 years vs 3·21 years; HR 0·38, 95% CI 0·31–0·47; ADT=androgen deprivation therapy. ECOG=Eastern Cooperative Oncology Group. PSA=prostate-specific antigen.
p<0·0001; table 2, appendix p 11). Among patients in SOC=standard of care. *The median number of docetaxel cycles was 6 (IQR 6–6) in both the SOC with abiraterone and
SOC without abiraterone groups. †Without visceral metastases. ‡With or without lymph node and bone metastases.
the ADT with docetaxel population who developed
§The metastatic burden was classified as reported by Sweeney and colleagues (2015), with a high burden characterised
castration resistance and were alive at the data cutoff by four or more bone metastases with one or more outside the vertebral bodies or pelvis, or visceral metastases, or
date, 221 (84%) of 263 patients were subsequently both; all other assessable situations were classified as low burden.
treated by at least one life-prolonging therapy and
Table 1: Baseline characteristics in the intention-to-treat population
213 (81%) of 263 by at least one next-generation
hormonal therapy in the SOC without abiraterone
groups, compared with 104 (74%) of 141 and 65 (46%) the addition of docetaxel to abiraterone did not increase
of 141, respectively, in the SOC plus abiraterone groups the incidence of severe or fatal adverse events. In the
(appendix p 14). Prostate-cancer-specific survival was ADT with docetaxel population, 49 severe adverse events
also improved by the addition of abiraterone in the ADT occurred per 100 person-years in patients who received
with docetaxel population (median 4·72 years vs not abiraterone compared with 55 per 100 person-years in
reached; HR 0·69, 95% CI 0·53–0·90; p=0·0062; patients who did not receive abiraterone (appendix p 14).
table 2, appendix p 11). The incidence of frequent severe adverse events (grade 3
Adding abiraterone to ADT with docetaxel (with or or worse adverse events that were reported in ≥5% of
without radiotherapy) had no effect on the number of patients) was similar between patients who received
docetaxel cycles administered (median 6 [IQR 6–6] in abiraterone and patients who did not in the ADT with
both groups). 138 (61%) of 226 patients in the ADT with docetaxel population, except for hypertension (76 [22%]
docetaxel population and 183 (53%) of 347 in the ADT of 347 patients vs 45 [13%] of 350) and hepatotoxicity with
without docetaxel population had abiraterone treatment increased aminotransferases (20 [6%] of 347 vs two [1%]
discontinued, including 29 (21%) of 138 and 32 (17%) of of 350), which were more frequent among those who
183 due to toxicity, respectively (appendix p 14). Overall, received abiraterone (table 3). Although the incidence of
the median abiraterone treatment duration before severe hypertension and hepatotoxicity were similar in
discontinuation was 33·2 months (95% CI 25·5–43·2) in patients who received abiraterone in the ADT with
the ADT without docetaxel population and 34·1 months docetaxel and ADT without docetaxel populations, there
(30·0–43·5) in the ADT with docetaxel population. was a difference, as expected, in the incidence of severe
In the ADT with docetaxel safety population, 217 (63%) neutropenia, which occurred in 34 (10%) of 347 patients
of 347 patients who received abiraterone versus 181 (52%) versus none of 226, respectively (appendix p 15).
of 350 who did not receive abiraterone had at least Although severe fatigue and peripheral neuropathy
one severe adverse event (grade 3 or worse) and events did not reach the frequency cutoff of 5% in the
seven (2%) versus three (1%) had a fatal adverse event ADT with docetaxel population, ten (3%) of 347 patients
(table 3, appendix p 14). In the ADT without docetaxel who received abiraterone and 15 (4%) of 350 who did
population, 149 (66%) of 226 patients who received not receive abiraterone had at least one event of severe
abiraterone had at least one severe adverse event and fatigue, and four (1%) and six (2%) had at least one
eight (4%) had a fatal adverse event (table 2). Therefore, severe peripheral neuropathy event (table 3).

www.thelancet.com Vol 399 April 30, 2022 1701

 Articles

Discussion Whether the systemic treatment of high-volume disease

To our knowledge, this is the first trial to show that a (ie, extended metastatic spread) should differ from that of
triple systemic therapy, consisting of ADT, docetaxel, low-volume disease has been extensively debated. The
and a second-generation androgen signalling inhibitor benefit of ADT with docetaxel has been clearly shown for
(abiraterone), extends radiographic progression-free survival patients with mCSPC with high-volume disease,4,22 but
and overall survival in patients with de novo mCSPC. data collected from larger populations support a similar
benefit across high-volume and low-volume disease.3 This
discrepancy might result from different ratios of patients
SOC without abiraterone groups
SOC plus abiraterone groups
with de novo versus relapsing low-volume disease being
A Overall population B ADT with docetaxel population
included in these trials, because recurrences favour better
100 prognoses.23 Second-generation androgen receptor axis
progression-free survival (%)

inhibitors have shown overall survival benefits in patients

80
with mCSPC regardless of the disease volume
Radiographic

60 considered.5,9,10 In this trial, the mature data for men with

40 high-volume disease indicate a clear improvement in
both radiographic progression-free survival and overall
20
HR 0·54 (99·9% CI 0·41–0·71); p<0·0001 HR 0·50 (99·9% CI 0·34–0·71); p<0·0001 survival when abiraterone is added to ADT with docetaxel.
0 Even though the data on overall survival were immature
0 1 2 3 4 5 6 0 1 2 3 4 5
Number at risk
for patients with low-volume disease, an abiraterone-
SOC without 589 453 274 158 72 31 7 355 274 137 61 16 0 mediated benefit in radiographic progression-free
abiraterone groups survival was already apparent. Thus, ADT with docetaxel
SOC plus 583 495 355 230 119 47 12 355 303 200 105 35 0
abiraterone groups plus abiraterone should be proposed as a new SOC for
patients presenting with high-volume disease who are fit
C Overall population D ADT with docetaxel population
for docetaxel prescription. However, for patients with low-
100
volume disease, the decision of whether to use two or
80 three agents should be carefully considered by both the
Overall survival (%)

60 patients and their oncologists, to determine whether a

major benefit in radiographic progression-free survival
40
(HR 0·58) is sufficient to justify a treatment inten­
20
HR 0·82 (95·1% CI 0·69–0·98); p=0·030 HR 0·75 (95·1% CI 0·59–0·95); p=0·017
sification in a context of immature overall survival data. It
0 would also depend on whether an overall survival benefit
0 1 2 3 4 5 6 0 1 2 3 4 5 is required. Long-term data and consensus conferences
Number at risk
SOC without 589 556 480 334 207 101 37 355 329 281 172 78 18
will allow for clinical practice guidance refinement.
abiraterone groups As abiraterone can inhibit CYP3A4 in vitro,24 and as
SOC plus 583 541 470 340 230 111 47 355 328 287 183 98 25 CYP3A4 is involved in docetaxel elimination,25 the
abiraterone groups
concomitant administration of these two drugs could
E ADT with docetaxel population with F ADT with docetaxel population with potentially increase docetaxel exposure and thereby its
low-volume metastatic burden high-volume metastatic burden toxicity. However, no increase in chemotherapy-related
100
side-effects occurred in reaction to this concomitance,
80 which might suggest an absence of substantial drug–
Overall survival (%)

60 drug interactions between abiraterone and docetaxel.

This finding is consistent with the observation that
40
abiraterone and docetaxel have similar pharmacokinetic
20 behaviour when administered alone or in combination.26
HR 0·83 (95·1% CI 0·50–1·39); p=0·66 HR 0·72 (95·1% CI 0·55–0·95); p=0·019
0 Similarly, abiraterone did not affect the clearance of
0 1 2 3 4 5 0 1 2 3 4 5 another taxane, cabazitaxel, by CYP3A.27 We also did not
Time since randomisation (years) Time since randomisation (years)
Number at risk observe an increase in haematological toxicity when
SOC without 123 119 110 71 39 12 232 210 171 101 39 6 abiraterone was added to ADT with docetaxel, contrary to
abiraterone groups
SOC plus 131 127 116 80 41 9 224 201 171 103 57 16 a phase 2 trial in patients with mCRPC.28 The most
abiraterone groups frequent high-grade adverse event reported among
patients who received abiraterone in this trial was
Figure 2: Kaplan-Meier estimates of radiographic progression-free survival and overall survival in the overall
population and ADT with docetaxel population hypertension (22% of patients who received abiraterone
Time-to-event curves are presented for radiographic progression-free survival (A) and overall survival (C) in the vs 13% of those who did not). This expected adverse event
overall population, radiographic progression-free survival (B) and overall survival (D) in the ADT with docetaxel was carefully monitored and managed to prevent the
population, and overall survival in the ADT with docetaxel population in patients with low-volume metastatic
onset of cardiovascular events, which were not seen in
burden (E) and high-volume metastatic burden (F). SOC in the overall population was ADT with or without
docetaxel. SOC in the ADT with docetaxel population was ADT with docetaxel. ADT=androgen deprivation therapy. excess in this trial. In the ENZAMET trial,10 hypertension
SOC=standard of care (with or without radiotherapy). was also two times as frequent among patients with

1702 www.thelancet.com Vol 399 April 30, 2022

 Articles

Patients assessed, n Median, years Median difference, years Hazard ratio p value
SOC with SOC without SOC with SOC without
abiraterone abiraterone abiraterone abiraterone
groups groups groups groups
Primary outcomes in the overall population
Overall survival 583 589 5·7 4·7 0·9 (95·1% CI 0·0–2·0) 0·82 (95·1% CI 0·69–0·98) 0·030
Radiographic progression- 583 589 4·5 2·2 2·1 (99·9% CI 0·7–2·9) 0·54 (99·9% CI 0·41–0·71) <0·0001
free survival
Secondary outcomes in the overall population
CRPC-free survival 583 589 3·8 1·5 2·3 (95% CI 1·6–3·0) 0·40 (95% CI 0·35–0·47) <0·0001
Prostate-cancer-specific 583 589 NR 5·8 NA 0·75 (95% CI 0·61–0·91) 0·0038
survival
Primary outcomes in the ADT with docetaxel population
Overall survival 355 355 NR 4·4 NA 0·75 (95·1% CI 0·59–0·95) 0·017
Radiographic progression- 355 355 4·5 2·0 2·2 (99·9% CI 0·6–2·8) 0·50 (99·9% CI 0·34–0·71) <0·0001
free survival
Secondary outcomes in the ADT with docetaxel population
Overall survival in patients 131 123 NR NR NA 0·83 (95·1% CI 0·50–1·39) 0·66
with low-volume
metastatic burden
Overall survival in patients 224 232 5·1 3·5 1·1 (95·1% CI 0·2–1·9) 0·72 (95·1% CI 0·55–0·95) 0·019
with high-volume
metastatic burden
Radiographic progression- 129 122 NR 2·7 NA 0·58 (99·9% CI 0·29–1·15) 0·0061
free survival in patients
with low-volume
metastatic burden
Radiographic progression- 225 231 4·1 1·6 2·2 (99·9% CI 0·6–3·2) 0·47 (99·9% CI 0·30–0·72) <0·0001
free survival in patients
with high-volume
metastatic burden
CRPC-free survival 355 355 3·2 1·4 2·0 (95% CI 1·5–3·1) 0·38 (95% CI 0·31–0·47) <0·0001
Prostate-cancer-specific 355 355 NR 4·7 NA 0·69 (95% CI 0·53–0·90) 0·0062
survival
ADT=androgen deprivation therapy. CRPC=castration-resistant prostate cancer. NA=not available. NR=not reached. SOC=standard of care (with or without radiotherapy).

Table 2: Efficacy outcomes in the intention-to-treat population

mCSPC who received enzalutamide in addition to ADT Aside from the limitations inherently associated with
with docetaxel compared with those who did not receive open-label trials, the rapid clinical practice evolution
enzalutamide. However, with enzalutamide addition, for mCSPC that took place during the conduct of
higher incidences of fatigue, syncope, and therapy this trial prompted the sequential amendment of the
discontinuation were reported, whereas we did not PEACE-1 protocol and statistical analysis plan to reflect
observe a rise in such events in PEACE-1 with the addition those changes, answer meaningful questions, and
of abiraterone to ADT with docetaxel. Aminotransferase implement the best treatment options to all participants
increase, a common side-effect of abiraterone, was (appendix pp 7–8). In particular, as an overall survival
observed in this trial (6% of patients who received benefit for patients with mCSPC treated with abiraterone
abiraterone vs 1% of those who did not), but occurred at a combined with ADT alone was shown in 2017, accrual was
similar frequency with or without docetaxel. This then restricted to patients receiving ADT with docetaxel as
aminotransferase increase had no detectable clinical SOC, and our pre-established statistical analysis was
consequences, as recommended management was revised to specifically include the assessment of the
followed.29 Notably, the addition of abiraterone to ADT efficacy of abiraterone in terms of survival outcomes in
with docetaxel did not affect severe neutropenia incidence the ADT with docetaxel population. De facto, further
(10% of patients who received abiraterone vs 9% of those allocating men to ADT only would have been unethical,
who did not). The impact of toxicity on participants will given that the ADT plus abiraterone combination was
be better understood after the formal analysis of quality- already known to improve overall survival.
of-life data, at which time the triplet therapy oncological Our findings show that combining abiraterone with
benefits might still outweigh the adverse effect of the ADT and docetaxel improves both radiographic
additional toxicity. progression-free survival and overall survival compared

www.thelancet.com Vol 399 April 30, 2022 1703

 Articles

A B
SOC plus SOC without Radiographic p value SOC plus SOC without Radiographic
abiraterone abiraterone progression-free abiraterone abiraterone progression-free p value
groups, n/N groups, n/N survival HR groups, n/N groups, n/N survival HR
(99·9% CI) (99·9% CI)

Radiotherapy 0·64 0·94

No 0·56 (0·38–0·82) 0·49 (0·30–0·82)
Yes 0·52 (0·35–0·77) 0·50 (0·30–0·84)
Docetaxel 0·17 NA
No (not yet SOC) 69/134 103/137 0·52 (0·31–0·87)
No (physician decision) 42/92 56/96 0·77 (0·39–1·51)
Yes (as SOC) 138/354 211/353 0·50 (0·35–0·72)
ECOG performance status 0·22 0·98
0 163/410 253/410 0·50 (0·36–0·70) 90/249 140/245 0·50 (0·32–0·78)
1–2 86/170 117/176 0·63 (0·39–1·00) 48/105 71/108 0·50 (0·27–0·93)
ADT type 0·16 0·30
GnRH agonist 154/391 233/391 0·53 (0·38–0·75) 81/219 120/221 0·53 (0·33–0·86)
GnRH antagonist 94/187 135/193 0·57 (0·37–0·89) 56/133 90/131 0·46 (0·26–0·80)
Surgical castration 1/2 2/2 0·06 (0·00–3·20) 1/2 1/1 0·07 (0·00–7·45)
Metastatic burden 0·41 0·38
High 157/330 234/334 0·51 (0·37–0·72) 97/225 156/231 0·47 (0·30–0·72)
Low 92/250 136/252 0·59 (0·38–0·92) 41/129 55/122 0·58 (0·29–1·15)
Overall 249/580 370/586 0·54 (0·41–0·71) 138/354 211/353 0·50 (0·34–0·71)

0·0 0·5 1·0 1·5 0·0 0·5 1·0 1·5

Favours SOC Favours SOC without Favours SOC Favours SOC without
plus abiraterone abiraterone plus abiraterone abiraterone

C D
SOC plus SOC without Overall p value SOC plus SOC without Overall p value
abiraterone abiraterone survival HR abiraterone abiraterone survival HR
groups, n/N groups, n/N (95·1% CI) groups, n/N groups, n/N (95·1% CI)

Radiotherapy 0·86 0·85

No 0·84 (0·65–1·08) 0·73 (0·52–1·03)
Yes 0·81 (0·63–1·04) 0·76 (0·54–1·08)
Docetaxel 0·33 NA
No (not yet SOC) 71/135 81/138 0·85 (0·61–1·17)
No (physician decision) 36/93 36/96 1·11 (0·70–1·76)
Yes (as SOC) 121/355 151/355 0·75 (0·59–0·95) 0·93
ECOG performance status 0·39
0 144/412 173/412 0·78 (0·62–0·97) 76/250 93/246 0·75 (0·56–1·02)
1–2 84/171 95/177 0·91 (0·68–1·22) 45/105 58/109 0·74 (0·50–1·09)
ADT type 0·52 0·98
GnRH agonist 144/392 171/392 0·77 (0·62–0·97) 72/219 88/222 0·76 (0·56–1·04)
GnRH antagonist 82/189 95/195 0·93 (0·69–1·25) 47/134 62/132 0·73 (0·50–1·06)
Surgical castration 2/2 2/2 0·46 (0·06–3·31) 2/2 1/1 0·71 (0·06–8·00)
Metastatic burden 0·33 0·64
High 152/331 183/336 0·77 (0·62–0·96) 92/224 120/232 0·72 (0·55–0·95)
Low 76/252 85/253 0·93 (0·69–1·28) 29/131 31/123 0·83 (0·50–1·39)
Overall 228/583 268/589 0·82 (0·69–0·98) 121/355 151/355 0·75 (0·59–0·95)

0·0 0·5 1·0 1·5 0·0 0·5 1·0 1·5

Favours SOC Favours SOC without Favours SOC Favours SOC without
plus abiraterone abiraterone plus abiraterone abiraterone

Figure 3: HRs for radiographic progression-free survival and overall survival by predefined stratification factors in the overall population and ADT with docetaxel population
(A) Radiographic progression-free survival in the overall population. (B) Radiographic progression-free survival in the ADT with docetaxel population. (C) Overall survival in the overall population.
(D) Overall survival in the ADT with docetaxel population. n/N represents events/patients. The number of events per patients in the radiotherapy subgroup analysis is not presented as the efficacy of
radiotherapy is still under investigation. HRs are plotted on a linear scale. At time of the radiographic progression-free survival analysis, metastatic burden data were not available for six patients who
were excluded from the Cox model. Overall survival analysis by predefined stratification factors was conducted in the intention-to-treat population. ADT=androgen deprivation therapy. ECOG=Eastern
Cooperative Oncology Group. GnRH=gonadotropin releasing hormone. HR=hazard ratio. NA=not applicable. SOC=standard of care (with or without radiotherapy).

1704 www.thelancet.com Vol 399 April 30, 2022

 Articles

with ADT and docetaxel without abiraterone. Indeed, ADT with docetaxel population ADT without docetaxel population
docetaxel and abiraterone have distinct mechanisms of
SOC plus SOC without SOC plus SOC without
action. In addition, they do not display absolute cross- abiraterone groups abiraterone groups abiraterone groups abiraterone groups
resistance, as abiraterone improves overall survival (with or without (with or without (with or without (with or without
after docetaxel treatment and docetaxel retains some radiotherapy; radiotherapy; radiotherapy; radiotherapy;
n=347) n=350) n=226) n=237)
activity after abiraterone failure.30,31 Collectively, these
observations provide a rationale for the survival benefits Any adverse events 346 (100%) 349 (100%) 226 (100%) 233 (99%)
resulting from their combination in this trial. However, Severe (grade ≥3) 217 (63%) 181 (52%) 149 (66%) 97 (41%)
adverse events
this study did not answer whether this triple
Fatal (grade 5) adverse 7 (2%) 3 (1%) 8 (4%) 5 (2%)
combination yields clinical advantages over ADT plus a events
second-generation androgen receptor axis inhibitor.
Frequent severe adverse events
Indeed, at the time when this trial was conducted
Hypertension 76 (22%) 45 (13%) 66 (29%) 38 (16%)
(2013–18), combining ADT with such an inhibitor had
Neutropenia 34 (10%) 32 (9%) 0 0
not been approved for mCSPC, thereby precluding
Hepatotoxicity 20 (6%) 2 (1%) 14 (6%) 3 (1%)
such a comparison within the trial. Incidentally, the
Febrile neutropenia 18 (5%) 19 (5%) 2 (1%) 1 (<1%)
ongoing phase 3 ARASENS (NCT02799602) and
Gamma-glutamyl 17 (5%) 14 (4%) 6 (3%) 4 (2%)
ENZAMET (NCT02446405) trials are also testing the transferase increase
combined efficacy of second-generation androgen Erectile dysfunction 7 (2%) 5 (1%) 12 (5%) 13 (5%)
receptor axis inhibitors (daroluta­mide and enzalutamide, Blood alkaline 15 (4%) 12 (3%) 6 (3%) 13 (5%)
respectively) with ADT and docetaxel, and their phosphatase increase
upcoming overall survival results will put the benefits Other severe adverse events
of abiraterone in perspective. Fatigue 10 (3%) 15 (4%) 3 (1%) 0
Reaching treatment synergy and preventing acquired Peripheral 4 (1%) 6 (2%) 1 (<1%) 0
resistance despite tumour clonal heterogeneity are neuropathy
cardinal aspects of mCSPC management that require Data are n (%). As the patients were not randomly assigned according to docetaxel prescription, toxicities recorded in
an optimisation of the treatment framework. In the ADT without docetaxel and ADT with docetaxel populations are not directly comparable. Percentages are rounded
PEACE-1, significant survival benefits were obtained to the nearest integer. The safety population includes patients who actually received the assigned treatment. Severe
adverse events (grade ≥3) were considered frequent if they occurred in at least 5% of patients in either group and
when abiraterone was added to ADT with docetaxel. are reported in decreasing order of occurrence according to the Medical Dictionary for Regulatory Affairs Preferred
Remarkably, these benefits were observed even though Term classification. ADT=androgen deprivation therapy. SOC=standard of care.
more than 80% of the SOC without abiraterone group
Table 3: Adverse events in the safety population
eventually received a second-generation androgen
receptor axis inhibitor (mostly abiraterone or enzaluta­
mide) to address cancer progression. Indeed, by contrast whether combining such an intensive first line systemic
with previous trials, PEACE-1 investigators were allowed treatment with radiotherapy to the primary tumour
to prescribe early salvage treatments to the control might provide further clinical benefits for patients with
group patients who showed evidence of cancer mCSPC. This upcoming analysis will be performed
progression. To this end, a PSA minimum value of just when the preplanned number of radiographic
0·50 ng/mL was used to define mCRPC, even in the progression-free survival and overall survival events is
absence of radiographic progression. Collectively, these reached in the population of men presenting with
observations suggest that the upfront combination low-volume metastatic dissemination.
implemented in this trial is superior to the sequential Contributors
administration of these (or similar) agents as occurred KF, AB, and SF conceived and designed the study. KF, JC, GR, RM, AF,
in the control group. Of note, abiraterone is approved BT, SS, DB, PR, GK, GG, FC, J-FB, AH, MS, AT-V, IL, LM, BL, SA-L, EM,
CEK, AE, AR, NM, FS, FP, M-EC-F, SVF, MJ, and AB were investigators
for metastatic prostate cancer in many countries and is who conducted the study and collected data. KF, IR, and SF directly
close to becoming a worldwide generic drug. accessed and verified the underlying data. KF, JC, GR, RM, DB, FB, BT,
In conclusion, this trial not only confirms that GK, GG, AT-V, and SF contributed to data analysis or data interpretation.
combining abiraterone with SOC improves outcomes in All authors contributed to data interpretation and writing, review, and
approval of the manuscript for submission.
men with de novo mCSPC, but it shows that combining
abiraterone with ADT and docetaxel improves overall Declaration of interests
KF reports consulting fees from Amgen, AstraZeneca, Astellas, Bayer,
and radiographic progression-free survival and other CureVac, Janssen, Novartis, Orion, Pfizer, and Sanofi; honoraria from
efficacy endpoints with little toxicity increase. AstraZeneca, Astellas, Bayer, Janssen, Novartis, and Sanofi;
Nonetheless, as PEACE-1 only included patients with and participation on a Data Safety Monitoring Board for Lilly.
de novo mCSPC, it is still unclear whether this triple JC reports grants from AB Science, Aragon Pharmaceuticals,
Arog Pharmaceuticals, Astellas Pharma, AstraZeneca,
therapy might benefit patients with metachronous Aveo Pharmaceuticals, Bayer, Blueprint Medicines Corporation,
mCSPC. Our findings cannot directly address whether BN Immunotherapeutics, Boehringer Ingelheim, Esperia, Bristol
this triplet systemic combination is superior to ADT Myers Squibb, Clovis Oncology, Cougar Biotechnology, Deciphera
and abiraterone. Longer follow-up is required to answer Pharmaceuticals, Exelixis, F Hoffmann-La Roche, Genentech,

www.thelancet.com Vol 399 April 30, 2022 1705

 Articles

GlaxoSmithKline, Incyte Corporation, Janssen-Cilag, Karyopharm References

Therapeutics, Laboratoires Leurquin Mediolanum, Lilly, Medimmune, 1 Gravis G, Fizazi K, Joly F, et al. Androgen-deprivation therapy alone
Millennium Pharmaceuticals, Nanobiotix, Novartis Farmacéutica, or with docetaxel in non-castrate metastatic prostate cancer
Pfizer, Puma Biotechnology, Sanofi-Aventis, SFJ Pharma, and Teva (GETUG-AFU 15): a randomised, open-label, phase 3 trial.
Pharma; consulting fees from Astellas Pharma, AstraZeneca, Bayer, Lancet Oncol 2013; 14: 149–58.
Bristol Myers Squibb, Johnson & Johnson, MSD Oncology, Novartis, 2 Sweeney CJ, Chen YH, Carducci M, et al. Chemohormonal therapy
Pfizer, Roche, and Sanofi; honoraria from Bayer, Johnson & Johnson, in metastatic hormone-sensitive prostate cancer. N Engl J Med 2015;
373: 737–46.
and Astellas Pharma; payment for expert testimony from Astellas
Pharma, AstraZeneca, Bayer, Bristol Myers Squibb, 3 James ND, Sydes MR, Clarke NW, et al. Addition of docetaxel,
zoledronic acid, or both to first-line long-term hormone therapy in
Johnson & Johnson, MSD Oncology, Novartis, Pfizer, Roche, and
prostate cancer (STAMPEDE): survival results from an adaptive,
Sanofi; and participation on a Data Safety Monitoring Board or
multiarm, multistage, platform randomised controlled trial. Lancet
Advisory Board for Astellas Pharma, AstraZeneca, Bayer, Bristol Myers 2016; 387: 1163–77.
Squibb, Johnson & Johnson, MSD Oncology, Novartis, Pfizer, Roche,
4 Kyriakopoulos CE, Chen YH, Carducci MA, et al. Chemohormonal
and Sanofi. GR reports honoraria from Astellas Pharma, AstraZeneca, therapy in metastatic hormone-sensitive prostate cancer: long-term
Janssen, Ipsen, and Sanofi; and support for attending meetings from survival analysis of the randomized phase III E3805 CHAARTED
Janssen and Ipsen. RM reports grants from Bayer; honoraria from Trial. J Clin Oncol 2018; 36: 1080–87.
MSD; and participation on a Data Safety Monitoring Board or Advisory 5 Clarke NW, Ali A, Ingleby FC, et al. Addition of docetaxel
Board for Bristol Myers Squibb, Clovis, Janssen, MSD, and Pfizer. to hormonal therapy in low- and high-burden metastatic hormone
AF reports honoraria from Astellas Pharma, AstraZeneca, Janssen, sensitive prostate cancer: long-term survival results from the
Novartis, and Sanofi. BT reports grants from Ferring and Bayer; STAMPEDE trial. Ann Oncol 2019; 30: 1992–2003.
personal fees from Amgen, Astellas Pharma, Ferring, Bayer, Novartis, 6 Fizazi K, Tran N, Fein L, et al. Abiraterone plus prednisone in
Myovant, and Sanofi; and non-financial support from Astellas and metastatic, castration-sensitive prostate cancer. N Engl J Med 2017;
Janssen. SS reports grants from Astellas Pharma, AstraZeneca, and 377: 352–60.
Janssen; and consulting fees, honoraria, and receipt of equipment from 7 James ND, de Bono JS, Spears MR, et al. Abiraterone for prostate
Astellas Pharma, AstraZeneca, Bayer, Ipesen, Janssen, and Takeda. cancer not previously treated with hormone therapy. N Engl J Med
DB reports honoraria from Amgen, Astellas Pharma, Bayer, Janssen, 2017; 377: 338–51.
and Novartis. GK reports grants from Astellas Pharma and Janssen; 8 Fizazi K, Tran N, Fein L, et al. Abiraterone acetate plus prednisone
and participation on an Advisory Board and honoraria from Astellas in patients with newly diagnosed high-risk metastatic
Pharma, AstraZeneca, Janssen, and Sanofi. FC reports honoraria from castration-sensitive prostate cancer (LATITUDE): final overall
survival analysis of a randomised, double-blind, phase 3 trial.
Bristol Myers Squibb and MSD; and participation on an Advisory
Lancet Oncol 2019; 20: 686–700.
Board for AstraZeneca, Bristol Myers Squibb, MSD, Pfizer, and Ipsen.
9 Chi KN, Chowdhury S, Bjartell A, et al. Apalutamide in patients
AT-V reports grants from Bayer, Ipsen, and Pfizer; and consulting fees
with metastatic castration-sensitive prostate cancer: final survival
or honoraria from Astellas Pharma, AstraZeneca, Bristol Myers
analysis of the randomized, double-blind, phase III TITAN study.
Squibb, Ipsen, JNJ, Novartis, MSD, and Roche. BL reports honararia J Clin Oncol 2021; 39: 2294–303.
from Janssen; and support for attending meetings from Astellas
10 Davis ID, Martin AJ, Stockler MR, et al. Enzalutamide with
Pharma, Janssen, and Pfizer. AR reports honoraria from Astellas standard first-line therapy in metastatic prostate cancer.
Pharma, Bayer, and Janssen; and support for attending meetings from N Engl J Med 2019; 381: 121–31.
Astellas Pharma, Ipsen, and Janssen. NM reports honoraria from 11 Armstrong AJ, Iguchi T, Azad AA, et al. Final overall survival (OS)
Astellas Pharma, Bayer, Ipsen, and MSD. All other authors declare no analysis from ARCHES: a phase III, randomized, double-blind,
competing interests. placebo (PBO)-controlled study of enzalutamide (ENZA) +
androgen deprivation therapy (ADT) in men with metastatic
Data sharing
hormone-sensitive prostate cancer (mHSPC). ESMO Congress;
The data collected and analysed for the purpose of the present
Sept 18, 2021 (abstr LBA25).
manuscript are not immediately available due to ethical and legal
12 Parker CC, James ND, Brawley CD, et al. Radiotherapy to the primary
restrictions. However, Unicancer will grant access to all deidentified tumour for newly diagnosed, metastatic prostate cancer (STAMPEDE):
individual data underlying the published results upon written and a randomised controlled phase 3 trial. Lancet 2018; 392: 2353–66.
detailed request originating from all personnel involved in cancer 13 Rydzewska LHM, Burdett S, Vale CL, et al. Adding abiraterone to
research, sent to peace1@unicancer.fr. androgen deprivation therapy in men with metastatic hormone-
Acknowledgments sensitive prostate cancer: a systematic review and meta-analysis.
We deeply appreciate the guidance provided by the members of the Eur J Cancer 2017; 84: 88–101.
Steering Committee and the Independent Data Monitoring Committee 14 Vale CL, Fisher DJ, White IR, et al. What is the optimal systemic
and are thankful to all patients who agreed to participate treatment of men with metastatic, hormone-naive prostate cancer?
A STOPCAP systematic review and network meta-analysis.
in this study. We also appreciate the help provided by members of the
Ann Oncol 2018; 29: 1249–57.
Department of Biostatistics and Epidemiology at Gustave Roussy
15 Feyerabend S, Saad F, Li T, et al. Survival benefit, disease progression
(University of Paris-Saclay, Villejuif, France). Namely, we would like to
and quality-of-life outcomes of abiraterone acetate plus prednisone
thank Nathalie Bouvet-Forteau for setting up the randomisation process, versus docetaxel in metastatic hormone-sensitive prostate cancer:
as well as Matthieu Texier and Aurélie Bardet for providing interim a network meta-analysis. Eur J Cancer 2018; 103: 78–87.
safety statistical analyses to the Independent Data Monitoring 16 Parker C, Castro E, Fizazi K, et al. Prostate cancer: ESMO Clinical
Committee meetings, and Marie-Laure Tanguy and Agnès Laplanche for Practice Guidelines for diagnosis, treatment and follow-up.
their statistical inputs. We are also grateful for the help of Amel Ann Oncol 2020; 31: 1119–34.
Belhimeur from the Institut du Cancer de Montpellier (Montpellier, 17 Schaeffer E, Srinivas S, Antonarakis ES, et al. NCCN Guidelines
France) for timely data management, and the assistance of Hélène insights: prostate cancer, version 1.2021. J Natl Compr Canc Netw
Ribault, Muriel Habibian, Jean-Carlos Cascais, Steeve Castiglione, 2021; 19: 134–43.
Sadène Tezkratt, Dorsafe Fekiri, Tri Hahn Dung Doan, and Ange 18 Cornford P, van den Bergh RCN, Briers E, et al. EAU-EANM-
Grigbale regarding trial management, under the direction of Soazig ESTRO-ESUR-SIOG Guidelines on prostate cancer. Part II—2020
Nenan (Unicancer, Kremlin-Bicêtre, France). We thank the European update: treatment of relapsing and metastatic prostate cancer.
Organisation for Research and Treatment of Cancer. We express our Eur Urol 2021; 79: 263–82.
gratitude for the financial support provided by Janssen-Cilag, Ipsen, and 19 Vogl UM, Beer TM, Davis ID, et al. Lack of consensus identifies
Sanofi, and the French Government (grants PHRC-K-13-050 and important areas for future clinical research: Advanced Prostate
PRT-K20-116). We would also like to thank Sebastien Marion (Unicancer, Cancer Consensus Conference (APCCC) 2019 findings. Eur J Cancer
Kremlin-Bicêtre, France) for writing support and figure design. 2022; 160: 24–60.

1706 www.thelancet.com Vol 399 April 30, 2022

 Articles

20 Fizazi K, Abrahamsson PA, Ahlgren G, et al. Achievements and 27 Massard C, Mateo J, Loriot Y, et al. Phase I/II trial of cabazitaxel
perspectives in prostate cancer phase 3 trials from genitourinary plus abiraterone in patients with metastatic castration-resistant
research groups in Europe: introducing the Prostate Cancer prostate cancer (mCRPC) progressing after docetaxel and
Consortium in Europe. Eur Urol 2015; 67: 904–12. abiraterone. Ann Oncol 2017; 28: 90–95.
21 Scher HI, Halabi S, Tannock I, et al. Design and end points of 28 Climent Duran MA, Font A, Duran I, et al. Randomized phase II
clinical trials for patients with progressive prostate cancer and study of docetaxel (D) + abiraterone acetate (AA) versus D after
castrate levels of testosterone: recommendations of the Prostate disease progression to first-line AA in metastatic castration-resistant
Cancer Clinical Trials Working Group. J Clin Oncol 2008; 26: 1148–59. prostate cancer (mCRPC): ABIDO-SOGUG trial. J Clin Oncol 2020;
22 Gravis G, Boher J-M, Chen Y-H, et al. Burden of metastatic castrate 38 (suppl): 95.
naive prostate cancer patients, to identify men more likely to benefit 29 Colomba E, Marret G, Baciarello G, et al. Liver tests increase on
from early docetaxel: further analyses of CHAARTED and GETUG- abiraterone acetate in men with metastatic prostate cancer: natural
AFU15 studies. Eur Urol 2018; 73: 847–55. history, management and outcome. Eur J Cancer 2020; 129: 117–22.
23 Francini E, Gray KP, Xie W, et al. Time of metastatic disease 30 Fizazi K, Scher HI, Molina A, et al. Abiraterone acetate for
presentation and volume of disease are prognostic for metastatic treatment of metastatic castration-resistant prostate cancer: final
hormone sensitive prostate cancer (mHSPC). Prostate 2018; overall survival analysis of the COU-AA-301 randomised, double-
78: 889–95. blind, placebo-controlled phase 3 study. Lancet Oncol 2012;
24 Deb S, Chin MY, Adomat H, Tomlinson Guns ES. Abiraterone inhibits 13: 983–92.
1α,25-dihydroxyvitamin D3 metabolism by CYP3A4 in human liver 31 de Bono JS, Smith MR, Saad F, et al. Subsequent chemotherapy
and intestine in vitro. J Steroid Biochem Mol Biol 2014; 144: 50–58. and treatment patterns after abiraterone acetate in patients with
25 van Waterschoot RA, Lagas JS, Wagenaar E, et al. Absence of both metastatic castration-resistant prostate cancer: post hoc analysis of
cytochrome P450 3A and P-glycoprotein dramatically increases COU-AA-302. Eur Urol 2017; 71: 656–64.
docetaxel oral bioavailability and risk of intestinal toxicity. Canxs
2009; 69: 8996–9002.
26 Tagawa ST, Posadas EM, Bruce J, et al. Phase 1b study of
abiraterone acetate plus prednisone and docetaxel in patients with
metastatic castration-resistant prostate cancer. Eur Urol 2016;
70: 718–21.

www.thelancet.com Vol 399 April 30, 2022 1707