An Updated Analysis of The Survival Endpoints of Ascende-Rt: Clinical Investigation

ARTICLE IN PRESS
www.redjournal.org
CLINICAL INVESTIGATION
An Updated Analysis of the Survival Endpoints

of ASCENDE-RT
Justin Oh, MD,*,y Scott Tyldesley, MD,*,y Howard Pai, MD,z Michael McKenzie, MD,*,y Ross Halperin, MD,x
Graeme Duncan, MD,*,y Gerard Morton, MB,║ Mira Keyes, MD,*,y Jeremy Hamm, MSc,{ and
W. James Morris, MD*,y
*
Department of Radiation Oncology, BC Cancer − Vancouver, Vancouver, British Columbia, Canada; yDepartment of Surgery,
Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, Canada; zDepartment of Radiation Oncology, BC
Cancer − Victoria, Victoria, British Columbia, Canada; xDepartment of Radiation Oncology, BC Cancer − Kelowna, Kelowna, British
Columbia, Canada; ║Department of Radiation Oncology, University of Toronto, Sunnybrook Health Sciences Centre, Toronto,
Ontario, Canada; and {Department of Population Oncology, BC Cancer Agency, Vancouver, British Columbia, Canada
Received Feb 11, 2022; Accepted for publication Nov 1, 2022
Purpose: Using the primary endpoint of time to biochemical progression (TTP), Androgen Suppression Combined with
Elective Nodal and Dose Escalated Radiation Therapy (ASCENDE-RT) randomized National Comprehensive Cancer
Network patients with intermediate and high-risk prostate cancer to low-dose-rate brachytherapy boost (LDR-PB) or
dose-escalated external beam boost (DE-EBRT). Randomization to the LDR-PB arm resulted in a 2-fold reduction in
biochemical progression compared with the DE-EBRT group at a median follow-up of 6.5 years (P < .001). Herein, the
primary endpoint and secondary survival endpoints of the ASCENDE-RT trial are updated at a 10-year median follow-
up.
Methods: Patients were randomly assigned to either the LDR-PB or the DE-EBRT arm (1:1). All patients received 1 year of
androgen deprivation therapy and 46 Gy in 23 fractions of pelvic RT. Patients in the DE-EBRT arm received an additional 32
Gy in 16 fractions, and those in the LDR-PB arm received an 125I implant prescribed to a minimum peripheral dose of 115 Gy.
Two hundred patients were randomized to the DE-EBRT arm and 198 to the LDR-PB arm.
Results: The 10-year Kaplan-Meier TTP estimate was 85% § 5% for LDR-PB compared with 67% § 7% for DE-
EBRT (log rank P < .001). Ten-year time to distant metastasis (DM) was 88% § 5% for the LDR-PB arm and
86% § 6% for the DE-EBRT arm (P = .56). There were 117 (29%) deaths. Ten-year overall survival (OS) estimates
were 80% § 6% for the LDR-PB arm and 75% § 7% for the DE-EBRT arm (P = .51). There were 30 (8%)
patients who died of prostate cancer: 12 (6%) in the LDR-PB arm, including 2 treatment-related deaths, and 18
(9%) in the DE-EBRT arm.
Conclusions: Men randomized to the LDR-PB boost arm of the ASCENDE-RT trial continue to experience a large
advantage in TTP compared with those randomized to the DE-EBRT arm. ASCENDE-RT was not powered to detect
differences in its secondary survival endpoints (OS, DM, and time to prostate cancer−specific death) and none are
apparent. Ó 2022 Elsevier Inc. All rights reserved.
Corresponding author: Scott Tyldesley, MD; E-mail: styldesley@bccancer.bc. injections used in the trial. Without these grants, ASCENDE-RT would not
ca have been possible.
Sources of support: The British Columbia Cancer Agency received Data sharing statement: Research data are stored in an institutional
unrestricted educational grants from Oncura Corporation, a division of GE repository and will be shared upon request to the corresponding author.
Healthcare, which manufactured the model 6711 125Iodine RapidStrand Supplementary material associated with this article can be found, in the
sources used in the trial, and Sanofi-Aventis Canada, the maker of the online version, at doi:10.1016/j.ijrobp.2022.11.005.
Suprefact and Eligard luteinizing hormone-releasing hormone depot
Int J Radiation Oncol Biol Phys, Vol. 000, No. 00, pp. 1−10, 2022
0360-3016/$ - see front matter Ó 2022 Elsevier Inc. All rights reserved.
https://doi.org/10.1016/j.ijrobp.2022.11.005
Descargado para Anonymous User (n/a) en CES University de ClinicalKey.es por Elsevier en diciembre 29, 2022. Para uso personal
exclusivamente. No se permiten otros usos sin autorización. Copyright ©2022. Elsevier Inc. Todos los derechos reservados.
ARTICLE IN PRESS
2 Oh et al. International Journal of Radiation Oncology Biology Physics
Introduction registered with the National Cancer Institute database

(NCT00175396).
Before the publication of the Androgen Suppression Com-
bined with Elective Nodal and Dose Escalated Radiation
Therapy trial (ASCENDE-RT), there were no randomized Interventions
prospective data regarding the efficacy of combining exter-
nal beam radiation therapy (EBRT) with a low-dose-rate All patients in the trial were prescribed ADT for 1 year
brachytherapy (LDR-PB) boost in the treatment of prostate in total. ADT was administered every 3 months with
cancer compared with dose- escalated EBRT (DE-EBRT) either Buserelin acetate 9.45 mg SubQ (Suprefact Depot)
using a uniform regimen of androgen deprivation therapy or leuprolide acetate 22.5 mg SubQ (Eligard). They also
(ADT), even though such protocols had been widely used received concurrent oral antiandrogen with either fluta-
for more than 2 decades. With a total accrual of 398 mide 300 mg tid or bicalutamide 50 mg daily for a mini-
patients, randomization to LDR-PB boost was associated mum of 4 weeks beginning with the first injection. After
with a 2-fold risk reduction in biochemical failure compared 8 months of neoadjuvant ADT, 46 Gy in 23 fractions of
with DE-EBRT.1 Specifically, the 7-year Kaplan-Meier (K- pelvic RT was delivered using a 4-field technique encom-
M) time to first of biochemical, local, or distant progression passing the prostate, seminal vesicles, and regional
(TTP) in the LDR-PB arm was 86% compared with 75% in lymph nodes. Subsequently, men in the DE-EBRT arm
the DE-EBRT arm. Consistent with previous randomized received an additional 32 Gy in 16 fractions of EBRT to
trials assessing dose escalation, other endpoints analyzed in the prostate immediately afterward, while patients in the
ASCENDE-RT, including overall survival (OS), time to dis- LDR-PB arm received an 125I implant (Oncura model
tant metastasis (DM), and time to prostate cancer specific 6711 sources as RapidStrand, prescribed to a minimum
death (PCSM), did not differ statistically between the 2 peripheral dose of 115 Gy) 2 to 3 weeks after the com-
treatment arms.1 The current publication presents the pri- pletion of pelvic RT.
mary outcome of TTP and the secondary survival outcomes Please refer to the initial ASCENDE-RT publication for
of ASCENDE-RT at a median follow-up of 10 years. details regarding statistical power, a consort diagram, fol-
low-up procedures, randomization procedures, and protocol
violations. Prostate specific membrane antigen (PSMA) or
Axumin positron emission tomography scan were not rou-
Methods and Materials
tinely available for follow-up patients.1
Patient recruitment, eligibility, and

randomization Endpoints and analysis
National Comprehensive Cancer Network (NCCN) patients The primary endpoint was TTP defined as the time from
with high- and intermediate-risk prostate cancer with East- the first ADT injection to first progression, which in all
ern Cooperative Oncology Group 0 to 2 performance status cases was biochemical progression as per the Phoenix
were eligible for the trial. As per NCCN, high-risk prostate threshold of nadir PSA + 2 ng/mL. Secondary endpoints
cancer was defined as Gleason score (GS) 8 or higher, initial included OS, DM, and event free survival (EFS), defined
prostate-specific antigen (iPSA) greater than 20 ng/mL, or as the time from the first ADT injection to any recur-
clinical stage T3a or higher disease.2 Intermediate-risk pros- rence of cancer or death from any cause. Analysis of
tate cancer was defined as having at least 1 of the following: PCSM was based on a definition of prostate cancer death
T2b-T2c, iPSA 10 to 20 ng/mL, or GS 7.2 Before randomiza- that included all deaths, regardless of apparent cause, if
tion, patients with GS ≥8 or iPSA >20 ng/mL were required the death occurred after the initiation of systemic ther-
to have a computed tomography scan of the abdomen and apy for metastatic prostate cancer. Also included as pros-
pelvis and a bone scan to assess radiologic or scintigraphic tate cancer−specific deaths are 2 treatment- related
evidence of nodal or DM. Men with iPSA level >40 ng/mL, deaths in the LDR-PB arm who were not known to have
T-stage ≥T3b, previous transurethral resection of the pros- cancer recurrence at the time of death.
tate, prior RT to the pelvis, pre-ADT prostate volume >75 The purpose of the current analysis is to update the pri-
cm3, or ineligibility for anesthesia were excluded before ran- mary endpoint and secondary survival endpoints. However,
domization. After informed written consent, each man was the justification for the timing of this reanalysis was based
stratified by NCCN group and randomly assigned to the on a proposed hazard ratio for PCSM of 0.5 favoring the
treatment arm (1:1) using a computer-generated block ran- brachytherapy arm and all other-cause mortality being bal-
domization. Patients were accrued from November 2002 to anced. If this assumption were true, it was estimated that
December 2011. Two-hundred patients were randomized to there would be sufficient statistical power to identify such a
the DE-EBRT arm, and 198 patients were randomized to large effect on PCSM when 29 or more prostate- specific
the LDR boost arm. The study received approval from the deaths occurred in the setting of fewer than 99 deaths from
appropriate institutional research ethics boards and is all other causes in both arms. These conditions were
ARTICLE IN PRESS
Volume 00 Number 00 2022 ASCENDE-RT update 3
satisfied as of July 2019, and the data were locked down and competing risk analyses were performed and are included in
prepared for analysis. the Supplementary Materials (Figures E5-E8). Univariable
In addition, 4 post hoc analyses are included in this analysis (UVA) and multivariable analysis (MVA) were
update: undertaken using clinically relevant demographic and dis-
ease variables, including age, randomization arm, percent
1. As per Lo and Crook et al,3,4 the authors have included positive cores on biopsy (PPC), clinical T-stage, NCCN risk
an analysis of “biochemical cure,” namely the probability stratification, iPSA, number of high-risk features (T3a, iPSA
of having a residual PSA ≤0.2 ng/mL at 4-years follow- >20 ng/mL, Gleason score ≥8, or >50 of cores positive
up from the date of brachytherapy. Briefly, ASCENDE- [PPC]), and Gleason grade group (GGG), determined by
RT patients who had at least 4 years of follow-up from the worst Gleason core from the reviewing pathologist. Ran-
the date of treatment completion without biochemical domization arm and the variables with P value < .3 on UVA
progression as per Phoenix definition and who had a were included in the MVA (conditional, backward). Bino-
PSA value recorded between 3.5 to 4.5 years were strati- mial logistic regression analysis was performed to assess the
fied by whether they had PSA ≤0.2 at 4 years (or the factors associated with achieving PSA ≤0.2 at 4 years.
closest time point to 4 years). We also investigated As stratification arm did not satisfy proportionality
whether having a 4-year PSA >0.2 carried any implica- assumptions in the Cox model for the primary endpoint of
tion for longer term TTP. TTP, hazards are presented as both averaged hazard ratios
2. A second post hoc analysis assessed the subsequent use and time- dependent hazard ratios for completeness. Five-
of systemic therapy for the LDR-PB and DE-EBRT arms year time interval was chosen a priori based on a clinical
for a subset (n = 356) of patients randomized in rationale that the early biochemical relapses due to occult
ASCENDE-RT with long-term population-based phar- metastatic disease would tend to occur within 5 years, in
macy data available up to June 2018. contrast to the biochemical relapses due to local failure,
3. Time to local failure (TTLF) was not reported in our pre- which would typically occur in the 5- to 10-year interval. In
vious publication. Of note, the original protocol included the 10- to 15-year interval, there was less than 25% of the
an ultrasound-guided sextant biopsy at 34 months (26 original trial participants who remained at risk of biochemi-
months post-RT) to assess local control, but this provi- cal progression, limiting the interpretability.
sion was formally removed from the protocol 4 years
into accrual, because of the perception by some investi-
gators of an unacceptable risk of infection/bleeding after Results
such biopsies, which led to low compliance. Therefore,
the TTLF reported in this post hoc analysis was generally Primary endpoint (time to progression)
restricted to the local failures that were identified only at
digital rectal exam or imaging abnormality in follow-up In total, 95 patients had biochemical progression (Table 1).
and prompted a posttreatment biopsy. By randomization, the 10-year K-M TTP was 67% for the
4. EFS has been added to this analysis for direct comparison DE-EBRT arm compared with 85% for the LDR-PB arm
with other studies. EFS, unlike TTP, includes death from (Fig. 1; log rank P < .001). Within the NCCN intermediate-
any cause (as well as biochemical, local, or distant relapses) risk group, the 10-year K-M TTP was 73% for the DE-
as an event, with the patients censored at last follow-up EBRT arm compared with 90% for the LDR-PB arm (log
only if they were both alive and free of any recurrence. rank P = .02; Fig. E4A). Within the NCCN high-risk group,
the 10-year K-M TTP was 64% for the DE-EBRT arm com-
pared with 81% for the LDR-PB arm (log rank P = .006; Fig.
Statistical methods
E4B). In the MVA, randomization arm, PPC, T-stage, and
iPSA were associated with TTP (Table 2). The 10-year
K-M analyses were performed for TTP, OS, DM, TTLF, EFS, cumulative incidence of biochemical progression in the DE-
and PCSM to enable comparison with the prior publication. EBRT arm was 30% compared with 15% in the LDR-PB
For OS, TTLF, DM, and TTP, patients were censored at last arm (P = .0006; Fig. E5).
follow-up. Events in TTLF, DM, and TTP were local recur-
rence, metastatic recurrence, and any first progression
(which was always biochemical progression as per the Phoe- Overall metastatic relapse (nodal and/or distant
nix definition), respectively. No metastatic events occurred recurrence)
before local recurrence events in the TTLF analysis. For
PCSM, patients were censored at death from other cause or Combining both nodal and distant recurrences, 50 (13%)
last follow-up. Events in PCSM were deaths as defined pre- patients had metastatic relapse: 27 in the DE-EBRT (14%)
viously and included 2 treatment-related deaths in the LDR- and 23 (12%) in the LDR-PB arm by randomization
PB arm, and all patients who died of any other cause were (Table 1). Ten-year K-M DM was 86% and 88% for DE-
censored at last available follow-up before death. In addi- EBRT and LDR-PB, respectively (Fig. 2A; P = .56). On
tion, for TTP, DM, TTLF, and PCSM, Fine and Gray UVA, PPC, clinical T-stage, NCCN risk stratification,
ARTICLE IN PRESS
Table 1 Disease status at data lockdown (July 1, 2019) by randomization and actual treatment arm received
By randomization By actual treatment received
All patients DE-EBRT LDR-PB DE-EBRT LDR-PB Neither
Analysis (n = 398) (n = 200) (n = 198) (n = 195) (n = 188) (n = 15)
Relapsed* 95 (23.9) 63 (31.5) 32 (16.2) 60 (30.8) 28 (14.9) 7
Nonrelapsed 303 (76.1) 137 (68.5) 166 (83.8) 135 (69.2) 160 (85.1) 8
Metastatic disease 50 (12.6) 27 (13.5) 23 (11.6) 26 (13.3) 21 (11.2) 3
Alive 281 (70.1) 138 (69.0) 143 (72.2) 131 (67.2) 140 (74.5) 10
Deceased 117 (29.4) 62 (31.0) 55 (27.8) 64 (32.8) 48 (25.5) 5
ANED 221 (55.5) 99 (49.5) 122 (61.6) 94 (48.2) 121 (64.4) 6
DNED 83 (20.9) 39 (19.5) 44 (22.2) 42 (21.5) 39 (21.3) 2
AWD 60 (15.1) 39 (19.5) 21 (10.6) 37 (19.0) 19 (10.1) 4
DOWD 34 (8.5) 23 (11.5) 11 (5.5) 22 (11.3) 9 (4.8) 3
Died of prostate cancery 30 (7.5) 18 (9.0) 12 (6.1) 18 (9.2) 10 (5.3) 2
K-M estimates +/− 95% CI
TTP (%)
10 y 76 § 5 67 § 7 85 § 5 68 § 7 86 § 5
OS (%)
10 y 77 § 4 75 § 7 80 § 6 73 § 7 82 § 6
DM (%)
10 y 87 § 4 86 § 6 88 § 5 86 § 5 88 § 4
PCSM (%)
10 y 93 § 3 92 § 4 95 § 3 92 § 4 95 § 3
Abbreviations: ANED = alive, no evidence of disease; AWD = alive with disease; DE-EBRT = dose-escalated external beam radiation therapy; DM = time
to distant metastasis; DNED = dead, no evidence of disease; DOWD = dead of/with disease; K-M = Kaplan-Meier; LDR-PB = low-dose-rate prostate
brachytherapy; OS = overall survival; PCSM = time to prostate cancer specific death; TTP = time to progression.
*
TTP was defined as the absence of any biochemical (nadir prostate-specific antigen level plus 2 ng/mL threshold), imaging, or clinical recurrence of pros-
tate cancer and no receipt of any form of secondary treatment for prostate cancer after completion of protocol interventions.
y
Patients undergoing systemic treatment for metastatic prostate cancer at death were scored as having died of prostate cancer, regardless of the proximate
cause of death.
number of high-risk features, and GGG were associated with Time to prostate cancer−specific death
DM (Table E6). On MVA, PPC, clinical T-stage, and GGG
continued to remain significant (Table E6). The 10-year Using a definition established before unlocking the data for
cumulative incidence of metastatic failure in the DE-EBRT analysis, 30 trial subjects died of prostate cancer (Table 1): 18
arm was 13% compared with 12% in the LDR-PB arm were randomized to the DE-EBRT and 12 to the LDR-PB
(P = .61; Fig. E6). arm, which included 2 deaths that have been assigned to tox-
icity from treatment. One patient developed debilitating pelvic
pain that required inpatient management and died of an acute
OS myocardial infarction and pulmonary embolism, and another
patient died after receiving major pelvic surgery because of
In total, 117 patients have died. Of the 117 deaths, 34 men Fournier gangrene after brachytherapy. There were no deaths
died with evidence of recurrent prostate cancer; the remain- attributable to DE-EBRT toxicity. The 10-year K-M PCSM
ing 83 deaths occurred in men without evidence of recur- estimate was 93% for the entire study cohort. Analyzed by
rence (Table 1). Median survival was not reached, and the randomization, the 10-year K-M PCSM was 95% for the
10-year K-M OS estimate was 77%. OS did not differ signifi- LDR-PB arm versus 92% for the DE-EBRT arm (P = .26,
cantly between the arms, with 10-year K-M of 80% for figure not shown). Ten-year cumulative incidence of PCSM
LDR-PB versus 74% for DE-EBRT (P = .51; Fig. 2B). Only was 5.5% for the LDR-PB arm and 8.5% for the DE-EBRT
age at randomization and biochemical relapse status were arm (P = .286; Fig. E7). In MVA (Table E7), only PPC was
significantly associated with OS in MVA (Table 3). associated with PCSM. There were no statistical differences in
ARTICLE IN PRESS
Fig. 1. Time to biochemical progression (TTP) by randomization arm. Black: Dose-escalated external beam radiation ther-
apy (DE-EBRT). Gray: Low-dose rate prostate brachytherapy (LDR-PB).
Table 2 UVA and MVA for biochemical failure* (Cox model; backward: conditional)
UVA MVA
Variables
HR 95% CI P value HR 95% CI P value
y,z
Randomization arm (DE-EBRT vs LDR-PB) 2.12 1.39−3.25 < .001 2.05 1.33−3.14 .001
0-5 y 1.39 0.79−2.43 .26 1.33 0.76−2.34 .32
5-10 y 4.71 2.06−1.079 < .001 4.81 2.10−11.01 < .001
10-15 y 2.35 0.59−9.41 .23 2.37 0.59−9.55 .22
y
PPC (unit = 1%) 1.02 1.01−1.02 < .001 1.01 1.01−1.02 .001
y,z
Clinical T-stage (T3a vs T1-T2b) 2.05 1.37−3.08 < .001 2.09 1.38−3.16 < .001
Risk code (high vs intermediate)z,x 1.47 0.91−2.23 .11 NA NA NA
Log iPSA (unit = 1 log)y 2.58 1.18−5.66 .02 3.01 1.38−6.57 .006
z,x
Number of high-risk features (≥3 vs ≤2) 3.04 1.92−4.81 < .001 NA NA NA
z
Gleason grade group (4-5 vs 1-3) 1.12 0.74−1.69 .59 NA NA NA
Age (unit = 1 y) 0.99 0.97−1.02 .66 NA NA NA
Abbreviations: DE-EBRT = dose-escalated external beam radiation therapy; HR = hazard ratio; iPSA = initial prostate-specific antigen; LDR-PB = low-
dose-rate prostate brachytherapy; MVA = multivariable analysis; PPC = percent positive cores on biopsy; UVA = univariable analysis.
*
Biochemical failure was the first or concurrent sign of recurrence in all patients.
y
Entered into MVA model if univariate P < .3.
z
Categorical variable.
x
Composite variables not entered into MVA.
ARTICLE IN PRESS
Fig. 2. (A) Time to distant metastasis (DM). (B) Overall survival (OS) by randomization arm. Black: Dose-escalated external
beam radiation therapy (DE-EBRT). Gray: Low-dose rate prostate brachytherapy (LDR-PB).
the frequency of other causes of death such as intrapelvic or patients in the LDR-PB arm with PSA >0.2 at the 4-year fol-
extrapelvic malignancy or cardiac death. low-up. Of those who had a 4-year PSA ≤0.2, the subse-
quent 10-year TTP was 97% compared with 60% in those
who had PSA >0.2 at 4 years (P < .001; Fig. E8). On MVA
Post hoc analyses binomial logistic regression analysis, randomization arm
(DE-EBRT vs LDR-PB: odds ratio [OR], 13.32; 95% confi-
Three-hundred and twenty patients had at least 4 years of dence interval [CI], 7.22-24.58; P < .001), as well as log
follow-up, a PSA value measured between 3.5 and 4.5 years, iPSA (OR, 3.88; 95% CI, 1.32-11.43; P = .014) and age (OR,
and no evidence of biochemical failure to that point. In this 0.94; 95% CI, 0.91-0.98; P = .003) were associated with
subset, there were 94 patients in the DE-EBRT arm and 19 residual PSA >0.2 at 4 years (Table E4).
Table 3 UVA and MVA for overall survival (Cox model; backward: conditional)
UVA MVA
Variables
HR 95% CI P value HR 95% CI P value
Randomization arm (DE-EBRT vs LDR-PB)* 1.13 0.79−1.63 .51 1.04 0.72−1.50 .84
y
PPC (unit = 1%) 1.01 0.99−1.01 .09 1.01 0.99−1.01 .12
,y
Clinical T-stage (T3a vs T1-T2b)* 1.23 0.84−1.80 .29 1.11 0.75−1.66 .60
,z
Risk code (high vs intermediate)* 1.25 0.84−1.86 .27 NA NA NA
Log iPSA (unit = 1 log) 1.22 0.61−2.44 .58 NA NA NA
Number of high-risk features (≥3 vs ≤2)* ,z
1.34 0.82−2.19 .24 NA NA NA
Gleason grade group (1-3 vs 4-5)* 1.18 0.81−1.71 .39 NA NA NA
Age (unit = 1 y)y 1.07 1.04−1.10 < .001 1.07 1.04−1.10 < .001
Disease status (relapse vs no relapse)* ,y
2.18 1.44−3.30 < .001 2.15 1.41−3.29 < .001
Abbreviations: DE-EBRT = dose-escalated external beam radiation therapy; HR = hazard ratio; iPSA = initial prostate-specific antigen; LDR-PB = low-
dose-rate prostate brachytherapy; MVA = multivariable analysis; PPC = percent positive cores on biopsy; UVA = univariable analysis.
*
Categorical variable.
y
Entered into MVA model if univariate P < .3.
z
Composite variables not entered into MVA.
ARTICLE IN PRESS
Fig. 3. Event-free survival (EFS) by randomization arm. Black: Dose-escalated external beam radiation therapy (DE-EBRT).
Gray: Low-dose rate prostate brachytherapy (LDR-PB).
Three-hundred and fifty-six patients (177 in DE- Discussion

EBRT and 179 in LDR-PB arms) had population-based
pharmacy data available after the completion of protocol Radiation dose escalation has been shown to reduce the inci-
treatment. There were 48 (24%) patients in the DE- dence of biochemical relapse compared with conventional
EBRT arm and 26 (13%) patients in the LDR-PB who curative doses, particularly in men with unfavorable prog-
received ADT at relapse. Overall, there were 644 and nostic features.5-7 ASCENDE-RT, which began accrual in
468 prescriptions for systemic therapy (chemotherapy, 2002, took the next logical step and employed a rigorously
antiandrogen, or lutenizing hormone-releasing hormone designed randomized trial to compare the relative efficacy
agonist) for prostate cancer recurrence or progression in of 2 entirely different methods of achieving radiation dose
the DE-EBRT and LDR-PB arms, respectively. escalation. Like other dose-escalation trials conducted at the
There were 19 (4.8%) local prostate relapses recorded; 17 time, pragmatic considerations, as explained in our original
were in men who received DE-EBRT and 2 in men who publication, led to the choice of TTP as the primary end-
received LDR-PB. The apparent 10-year K-M TTLF was point. In analyzing for TTP, deaths from unrelated causes
99% in the LDR-PB arm compared with 91% in the DE- were not scored as events; reporting TTP this way is consis-
EBRT arm (log rank P < .001; figure not shown). Ten-year tent with our previous publication and reflects the probabil-
cumulative incidence of LF was 1.5% for the LDR-PB arm ity of biochemical control for long-term survivors. In this
and 7.1% for the DE-EBRT arm (P = .0006; Fig. E9). On update, the authors have also provided an analysis of EFS in
UVA, randomization arm and PPC were both statistically which deaths from any cause are considered events to allow
associated with TTLF, and both variables remained highly comparison with other contemporary studies.
significant in MVA (Table E5). In this update with median follow of 10 years, a clear
For EFS, there were no local, regional, or distant recur- advantage in TTP and EFS for those subjects randomized to
rences before biochemical failure. By randomization, 10- the LDR-PB arm persists. Although numbers are small at
year K-M EFS was 54% for the DE-EBRT arm compared 15 years, it is estimated that only 20% of those in the LDR-
with 69% for the LDR-PB arm (Fig. 3; log rank P < .001). PB arm will have experienced biochemical recurrence
ARTICLE IN PRESS
compared with 47% of the patients in the DE-EBRT arm at mortality.1 We proposed that the observed reduction in OS
that point. The TTP outcomes after LDR-PB boost in among relapsed subjects was principally caused by a rela-
ASCENDE-RT are especially noteworthy because more tively small number of trial participants with occult meta-
than two-thirds of the trial subjects (276/398) were high risk static disease that was present before registration and
(HiR) by NCCN criteria and nearly half (193/398) had 2 or unrelated to the treatment arm, which was thoroughly dis-
more HiR features, which included PPC ≥50%, GGG 4 to 5, cussed in the original ASCENDE-RT publication. Because
iPSA >20 ng/mL, or T3a stage. Even within HiR subset, these trial subjects would have been allocated evenly to each
fewer than 25% of the trial subjects assigned to the LDR-PB arm by the randomization process, OS was not correlated
arm had a biochemical relapse. with treatment arm despite the relatively large difference in
Limitations of the ASCENDE-RT trial were discussed in biochemical relapse rates observed. This interpretation is
the original publication and are not repeated in detail herein supported by the observed reduction in the HR comparing
but remain relevant to the current analysis. It is possible that the current and previously published analyses: the relative
changes in clinical practice, such as adjuvant rather than neo- risk of death among those with biochemical relapse was
adjuvant ADT, longer durations in ADT, changes in planning more than 6 times greater than those without relapse in the
and prescribing external RT, newer staging modalities (eg, original analysis (HR of 6.3; 95% CI, 3.6-10.9; P < .001)
multiparametric MRI and PSMA positron emission tomogra- compared with an HR of just 2.2 in the current analysis
phy), changes in image guidance, and intensification of ADT (95% CI, 1.4-3.3; P < .001) (Table 3). This reduction in HR
(such as an addition of abiraterone), would alter the differen- suggests that deaths from prerandomization occult meta-
ces observed if this trial were repeated. Similarly, changes in static disease no longer dominate all-cause mortality
brachytherapy practice and dose may alter the side effect pro- because competing causes of death have become more com-
file previously described. Recent meta-analyses have sug- mon among the aging trial participants. The observation
gested that adjuvant ADT may be superior to neoadjuvant supports the efforts to better identify men with occult meta-
ADT for DM in localized prostate cancer.8,9 In addition, the static disease with more sensitive investigations such as
Randomised Androgen Deprivation and Radiotherapy trial PSMA scan to triage treatment plans and establish appropri-
has suggested improvement in DM with 18 months com- ate future clinical trial protocols.14,15
pared with 6 months of ADT among those who received There were 19 (4.8%) documented local failures, which
high-dose-rate (HDR) boost, although it consisted of non- almost certainly underreports the true local failure rate
randomized radiation dosing, and younger men with higher because the plan for routine systematic biopsies at defined
risks were associated with HDR boost treatment.10 posttreatment intervals did not prove feasible (as described
ASCENDE-RT was not designed to assess the sequencing or in the Methods and Materials section), and the sensitivity of
duration of ADT, but each arm received the same ADT regi- digital rectal exam is poor in detecting persistent/recurrent
men to ensure that the assessment of the effect of LDR-PB local disease.16,17 Thus, for most of the patients, unless there
could be isolated as much as possible. Some of the rationales was locally aggressive recurrence or a clinical situation
for neoadjuvant ADT and the duration included in the origi- where local salvage was contemplated, posttreatment biop-
nal protocol, written more than 20 years ago, included early sies were not typically performed, therefore it is difficult to
results of Radiation Therapy Oncology Group (RTOG) 9413 accurately estimate the true rate of local relapse. However,
as well as surgical and preclinical literature that have sug- there remains a large disparity in the number of local fail-
gested the benefit of neoadjuvant ADT.11-13 A recent meta- ures, with only 2 recorded among the 188 subjects who
analysis by Kishan et al8 proposed that neoadjuvant ADT received LDR-PB boost compared with 17 among the 195
may not be necessary, although the study did not demon- men who received DE-EBRT. Furthermore, there were an
strate inferiority of prolonged neoadjuvant ADT duration, additional 26 patients with biochemical relapses (1/4 of all
and no trial included in the meta-analyses used neoadjuvant biochemical relapses and which occurred disproportionally
ADT as long as ASCENDE-RT.9 It is possible that the trial among those assigned to DE-EBRT) in which the site of
subjects reported here may not have required or benefited relapse could not be identified despite a bone scan and com-
from neoadjuvant ADT or that a similar or longer duration puted tomography of chest/abdomen/pelvis.
of concurrent and adjuvant ADT may have resulted in equiv- Absolute PSA cutoff and relative PSA rise have both been
alent or even superior outcomes, particularly for the DE- used as surrogate endpoints for recurrence after definitive
EBRT arm. Data in this regard are lacking, and prospective therapy in prostate cancer, mostly based on large retrospec-
randomized trials with prespecified duration and sequencing tive studies.18-20 A reanalysis of TTP in ASCENDE-RT
of ADT for both DE-EBRT and LDR-PB arms would be using an absolute PSA cutoff of >0.2 ng/mL showed an even
required to answer such questions. For TTLF post hoc analy- larger advantage for the LDR-PB boost arm.21 In this
sis, patients were not censored at time of biochemical or dis- update, we show that the probability of having a PSA >0.2
tant metastasis (although no metastatic events occurred at 4 years is over 10 times more likely for trial subjects ran-
before local or regional relapse), which may have resulted in domized to the DE-EBRT arm compared with those in the
a bias that affected the local recurrence rate described. LDR-PB arm. In turn, PSA values ≤0.2 at 4 years after defin-
As in the previous analysis, there was strong statistical itive RT are associated with a high probability of long-term
correlation between biochemical failure and all-cause biochemical control.3,4 Recent analysis by Noble et al22 has
ARTICLE IN PRESS
also externally validated this surrogate outcome for long- and entered mainstream treatment protocols.36-38 Although
term durable biochemical control, demonstrating that studies featuring these modalities seem promising, direct
patients who achieved PSA ≤0.2 at 4 years after brachyther- comparison to ASCENDE-RT is limited by the short follow-
apy had greater than 97% 10-year disease-free survival. up or different inclusion criteria. To this date, there has not
Among ASCENDE-RT participants, the 10-year TTP was been a randomized prospective study with as many high-risk
only 60% for trial subjects with a 4- year PSA >0.2 versus patients, as long of a follow-up, and as effective biochemical
97% for those with ≤ 0.2, an observation that appears to control as the ASCENDE-RT trial. It may be true that other
apply to both treatments. However, we acknowledge that modalities may demonstrate better therapeutic ratios in the
this PSA cut off has not been fully validated for DE-EBRT, future. However, it is premature to conclude that an LDR-PB
and residual PSA at 4 years may reflect residual “healthy” boost necessarily has similar efficacy to or worse adverse
prostate tissue that may or may not influence future prostate effects than HDR-PB or ultrahypofractionation via stereotac-
cancer outcomes. tic techniques until these modalities are compared directly in
Although TTP has not reliably served as a surrogate clinical trials in which the prospective collection of the pri-
marker for OS in the randomized trials to date, biochemical mary outcome, toxicity, and quality of life data are collected
relapse continues to have clinical relevance. In addition to using similar tools. Exploration of these modalities will hope-
being quantitative, objective, and minimally invasive, bio- fully provide more options for patients for the treatment of
chemical failure faithfully predicts eventual clinical recur- localized prostate cancer, incorporating disease characteristics
rence and often signals the introduction of further treatment, and patient preference based on efficacy and toxicity.
including potentially morbid salvage therapies or the reintro-
duction of ADT, which is associated with multiple short- and
long-term adverse effects. To date, there have been almost Conclusions
40% more prescriptions for systemic agents for prostate can-
cer relapse in the DE-EBRT arm of ASCENDE-RT.23-30 LDR-PB boost continues to demonstrate superior and dura-
Our results are similar to previously published brachyther- ble time to biochemical progression and EFS compared with
apy boost studies. Hoskin et al31 recently updated their analy- DE-EBRT at 10 years for patients with intermediate- and
sis of relapse-free survival (equivalent to EFS as reported high-risk prostate cancer. LDR-PB also results in substan-
here), defined as the time to biochemical recurrence, clinical tially lower posttreatment (residual) PSA values and may be
progression, or death, of the patients who had cT1-T3 pros- associated with less subsequent use of systemic therapy. The
tate cancer and received HDR brachytherapy boost versus important balancing issue of treatment-related toxicity has
EBRT boost in addition to 6 to 36 months of ADT. The been reported in detail in our previous publications.
patients who had HDR brachytherapy boost continued to Although ASCENDE-RT was not designed for prolonged
demonstrate relapse-free survival improvement at 12 years prospective toxicity analysis, differences in toxicity, whether
compared with the EBRT boost group, but there was no sta- transient or permanent, remain important considerations.
tistically significant difference in DM or OS.31 Another study Despite the differences observed, both treatment arms in
by Sathya et al demonstrated time to biochemical progression ASCENDE-RT demonstrated high and equivalent rates of
improvement in patients with cT2-T3 prostate cancer who OS, time to DM, and prostate cancer−specific death among
received EBRT and iridium implant boost compared with patients with unfavorable prognostic features.
those who received EBRT alone at a median follow-up of
8 years.32,33 Subsequent 14-year follow-up study did not dem-
onstrate a statistically significant difference between the 2 Disclosures
groups in PCSM, DM, or OS.32 However, there were substan-
tially more patients who died of other causes than prostate
cancer, and potential OS benefit in a younger patient cohort G.M. serves as the chair of Americal Society for Radiation
therefore could not be completely excluded.32 Oncology (ASTRO) genitourinary track and Royal College
Compared with those trial subjects who received DE-EBRT, Brachytherapy Area of Focused Competency (AFC) com-
the improvement in TTP enjoyed by the LDR-PB boost arm mittee; S.T. has received honoria less than $5000 over 3
came at a price in terms of the severity and frequency of years from Bayer, Tera Sera, Astella, and Janssen. Other
adverse urinary side effects and a statistically significant reduc- authors have none to disclose.
tion in some quality of life domains.34 There were 2 toxic
deaths among trial participants and both were in the LDR-PB
arm. As stated in our prior publication, it is possible, but not References
certain, that technical modifications may improve the thera-
peutic ratio of brachytherapy boosts, and such approaches 1. Morris WJ, Tyldesley S, Rodda S, et al. Androgen suppression com-
bined with elective nodal and dose escalated radiation therapy (the
should be investigated in future prospective studies.34,35
ASCENDE-RT trial): an analysis of survival Endpoints for a random-
Since the design of the ASCENDE-RT trial other radiation ized trial comparing a low-dose-rate brachytherapy boost to a dose-
treatment modalities, including HDR boost, salvage brachy- escalated external beam boost for high- and intermediate-risk prostate
therapy, focal external beam boost, and SABR have emerged cancer. Int J Radiat Oncol Biol Phys 2017;98(2):275-285.
ARTICLE IN PRESS
2. Mohler JL, Antonarakis ES, Armstrong AJ, et al. Prostate cancer, ver- 22. Noble DJ, Doyle E, Tramonti G, et al. Defining biochemical cure after
sion 2.2019, NCCN Clinical Practice Guidelines in Oncology. J Natl low dose rate prostate brachytherapy: External validation of 4-year
Compr Canc Netw 2019;17:479-505. prostate-specific antigen nadir as a predictor of 10- and 15-year dis-
3. Lo AC, Morris WJ, Lapointe V, et al. Prostate-specific antigen at 4 to 5 ease-free survival. Clin Oncol. (R Coll Radiol). 2022;34:42-49.
years after low-dose-rate prostate brachytherapy is a strong predictor 23. Bosco C, Bosnyak Z, Malmberg A, Adolfsson J, Keating NL, Van Hem-
of disease-free survival. Int J Radiat Oncol Biol Phys 2014;88(1):87-93. elrijck M. Quantifying observational evidence for risk of fatal and non-
4. Crook JM, Tang C, Thames H, et al. A biochemical definition of cure fatal cardiovascular disease following androgen deprivation therapy for
after brachytherapy for prostate cancer. Radiother Oncol 2020;149:64-69. prostate cancer: A meta-analysis. Eur Urol 2015;68:386-396.
5. Al-Mamgani A, van Putten WL, Heemsbergen WD, et al. Update of 24. Gay HA, Sanda MG, Liu J, et al. External beam radiation therapy or
Dutch multicenter dose-escalation trial of radiotherapy for localized brachytherapy with or without short-course neoadjuvant androgen
prostate cancer. Int J Radiat Oncol Biol Phys 2008;72:980-988. deprivation therapy: Results of a multicenter, prospective study of qual-
6. Dearnaley DP, Jovic G, Syndikus I, et al. Escalated-dose versus control-dose ity of life. Int J Radiat Oncol Biol Phys 2017;98:304-317.
conformal radiotherapy for prostate cancer: Long-term results from the 25. Keating NL, Liu’PH O’Malley AJ, Freedland SJ, Smith MR. Androgen-
MRC RT01 randomised controlled trial. Lancet Oncol 2014;15:464-473. deprivation therapy and diabetes control among diabetic men with
7. Michalski JM, Moughan J, Purdy J, et al. Effect of standard vs dose- prostate cancer. Eur Urol 2014;65:816-824.
escalated radiation therapy for patients with intermediate-risk prostate 26. Nguyen PL, Alibhai SM, Basaria S, et al. Adverse effects of androgen
cancer: The NRG Oncology RTOG 0126 randomized clinical trial. deprivation therapy and strategies to mitigate them. Eur Urol
JAMA Oncol 2018;4 e180039. 2015;67:825-836.
8. Kishan AU, Sun Y, Hartman H, et al. Androgen deprivation therapy 27. O’Farrell S, Garmo H, Holmberg L, Adolfsson J, Stattin P, Van Hemel-
use and duration with definitive radiotherapy for localised prostate rijck M. Risk and timing of cardiovascular disease after androgen-dep-
cancer: An individual patient data meta-analysis. Lancet Oncol rivation therapy in men with prostate cancer. J Clin Oncol
2022;23:304-316. 2015;33:1243-1251.
9. Spratt DE, Malone S, Roy S, et al. Prostate radiotherapy with adjuvant 28. Williams S. Surrogate endpoints in early prostate cancer research.
androgen deprivation therapy (ADT) improves metastasis-free survival Transl Androl Urol 2018;7:472-482.
compared to neoadjuvant ADT: An individual patient meta-analysis. J 29. Collette L, Burzykowski T, Schroder FH. Prostate-specific antigen
Clin Oncol 2021;39:136-144. (PSA) alone is not an appropriate surrogate marker of long-term
10. Joseph D, Denham JW, Steigler A, et al. Radiation dose escalation or therapeutic benefit in prostate cancer trials. Eur J Cancer
longer androgen suppression to prevent distant progression in men 2006;42:1344-1350.
with locally advanced prostate cancer: 10-year data from the TROG 30. Collette L, Burzykowski T, Carroll KJ, et al. Is prostate-specific antigen
03.04 RADAR trial. Int J Radiat Oncol Biol Phys 2020;106:693-702. a valid surrogate end point for survival in hormonally treated patients
11. Gleave ME, Goldenberg SL, Chin JL, et al. Randomized comparative with metastatic prostate cancer? Joint research of the European Orga-
study of 3 versus 8-month neoadjuvant hormonal therapy before radi- nisation for Research and Treatment of Cancer, the Limburgs Universi-
cal prostatectomy: Biochemical and pathological effects. J Urol tair Centrum, and AstraZeneca Pharmaceuticals. J Clin Oncol
2001;166:500-506 discussion 506-507. 2005;23:6139-6148.
12. Roach 3rd M, DeSilvio M, Lawton C, et al. Phase III trial comparing 31. Hoskin PJ, Rojas AM, Ostler PJ, Bryant L, Lowe GJ. Randomised trial
whole-pelvic versus prostate-only radiotherapy and neoadjuvant versus of external-beam radiotherapy alone or with high-dose-rate brachy-
adjuvant combined androgen suppression: Radiation Therapy Oncol- therapy for prostate cancer: Mature 12-year results. Radiother Oncol
ogy Group 9413. J Clin Oncol 2003;21:1904-1911. 2021;154:214-219.
13. Zietman AL, Prince EA, Nakfoor BM, Park JJ. Androgen deprivation 32. Dayes IS, Parpia S, Gilbert J, et al. Long-term results of a randomized trial
and radiation therapy: Sequencing studies using the Shionogi in vivo comparing iridium implant plus external beam radiation therapy with
tumor system. Int J Radiat Oncol Biol Phys 1997;38:1067-1070. external beam radiation therapy alone in node-negative locally advanced
14. Hofman MS. ProPSMA: A callout to the nuclear medicine community cancer of the prostate. Int J Radiat Oncol Biol Phys 2017;99:90-93.
to change practice with prospective, high-quality data. J Nucl Med 33. Sathya JR, Davis IR, Julian JA, et al. Randomized trial comparing irid-
2020;61:676-677. ium implant plus external-beam radiation therapy with external-beam
15. Hofman MS, Lawrentschuk N, Francis RJ, et al. Prostate-specific mem- radiation therapy alone in node-negative locally advanced cancer of
brane antigen PET-CT in patients with high-risk prostate cancer before the prostate. J Clin Oncol 2005;23:1192-1199.
curative-intent surgery or radiotherapy (proPSMA): A prospective, 34. Rodda S, Morris WJ, Hamm J, Duncan G. ASCENDE-RT: an Analysis
randomised, multicentre study. Lancet 2020;395:1208-1216. of health-related quality of life for a randomized trial comparing low-
16. Johnstone PA, McFarland JT, Riffenburgh RH, Amling CL. Efficacy of dose-rate brachytherapy boost with dose-escalated external beam boost
digital rectal examination after radiotherapy for prostate cancer. J Urol for high- and intermediate-risk prostate cancer. Int J Radiat Oncol Biol
2001;166:1684-1687. Phys 2017;98(3):581-589.
17. Smith DS, Catalona WJ. Interexaminer variability of digital rectal 35. Rodda S, Tyldesley S, Morris WJ, et al. ASCENDE-RT. an analysis of
examination in detecting prostate cancer. Urology 1995;45:70-74. treatment-related morbidity for a randomized trial comparing a low-
18. Stephenson AJ, Kattan MW, Eastham JA, et al. Defining biochemical dose-rate brachytherapy boost with a dose-escalated external beam
recurrence of prostate cancer after radical prostatectomy: A proposal boost for high- and intermediate-risk prostate cancer. Int J Radiat
for a standardized definition. J Clin Oncol 2006;24:3973-3978. Oncol Biol Phys 2017;98(2):286-295.
19. Stephenson AJ, Scardino PT, Eastham JA, et al. Preoperative nomo- 36. Crook J, Rodgers JP, Pisansky TM, et al. Salvage low-dose-rate prostate
gram predicting the 10-year probability of prostate cancer recurrence brachytherapy: Clinical outcomes of a phase 2 trial for local recurrence
after radical prostatectomy. J Natl Cancer Inst 2006;98:715-717. after external beam radiation therapy (NRG Oncology/RTOG 0526).
20. Toussi A, Stewart-Merrill SB, Boorjian SA, et al. Standardizing the defi- Int J Radiat Oncol Biol Phys 2022;112:1115-1122.
nition of biochemical recurrence after radical prostatectomy-what 37. Kerkmeijer LGW, Groen VH, Pos FJ, et al. Focal boost to the intrapro-
prostate specific antigen cut point best predicts a durable increase and static tumor in external beam radiotherapy for patients with localized
subsequent systemic progression? J Urol 2016;195:1754-1759. prostate cancer: Results from the FLAME randomized phase III trial. J
21. Morris WJ, Pickles T, Keyes M. Using a surgical prostate-specific anti- Clin Oncol 2021;39:787-796.
gen threshold of >0.2 ng/mL to define biochemical failure for interme- 38. Widmark A, Gunnlaugsson A, Beckman L, et al. Ultra-hypofractio-
diate- and high-risk prostate cancer patients treated with definitive nated versus conventionally fractionated radiotherapy for prostate can-
radiation therapy in the ASCENDE-RT randomized control trial. cer: 5-year outcomes of the HYPO-RT-PC randomised, non-
Brachytherapy 2018;17(6):837-844. inferiority, phase 3 trial. Lancet 2019;394:385-395.

An Updated Analysis of The Survival Endpoints of Ascende-Rt: Clinical Investigation

Uploaded by

Copyright:

Available Formats

An Updated Analysis of The Survival Endpoints of Ascende-Rt: Clinical Investigation

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

An Updated Analysis of The Survival Endpoints of Ascende-Rt: Clinical Investigation

Uploaded by

Copyright:

Available Formats

ARTICLE IN PRESS

An Updated Analysis of the Survival Endpoints

Received Feb 11, 2022; Accepted for publication Nov 1, 2022

Introduction registered with the National Cancer Institute database

Patient recruitment, eligibility, and

Three-hundred and ﬁfty-six patients (177 in DE- Discussion

You might also like