Evaluation of biologically equivalent dose escalation, clinical outcome, and toxicity in prostate cancer radiotherapy: A meta-analysis of 12,000 patients from 40 institutions Nicholas G Zaorsky, MD; Mark D Hurwitz, MD; Scott W Keith, PhD; Adam P Dicker, MD, PhD; Robert B Den, MD Department of Radiation Oncology Jefferson Medical College and Kimmel Cancer Center of Thomas Jefferson University Monday, 23 September 2013 10:45 AM

Disclosures None

Background The role of dose-escalation (DE) to a biologically equivalent dose (BED) of ~180 Gy (at an α/β of 1.5) with conventionally fractionated RT has been established to improve rates of freedom from biochemical failure (FFBF) Hypofractionated RT and high dose rate brachytherapy (HDR-BT) allow for further BED escalation

Background BED = (n) (d) [1 + ] (d) (α/β) [1 + ] (d) (α/β) n is the number of radiation fractions d is the fraction size (Gy) BED, at α/β of 1.5 = efficacy for eliminating prostate cancer 3 = late toxicity 10 = early toxicity

CFRT or HFRT, to a lower total dose (i.e. less fractions) SBRT “virtual HDR-BT” ~8 weeks 1.8 – 2.0 Gy 5 days per week 1 fraction, 15 minutes per day 5 days per week ~4 weeks 2.1 – 3.5 Gy 1 fraction, 15 minutes per day ~ 5 fractions 1 fraction, < 45 minutes per day 3.5 – 15.0 Gy ~1-2 weeks non-DE DE HDR-BT ~ 6 - 12 Gy per fraction ~ 1 - 5 days 1 fraction, ~2 hours per day CFRT or HFRT, to a lower total dose (i.e. less fractions) HDR-BT HDR-BT boost

n is the number of radiation fractions; Higher BED theorized to improve outcome Lower BED theorized to decrease toxicity BED = (nd[1 + d/(α/β)]) n is the number of radiation fractions; d is the fraction size

n is the number of radiation fractions; Higher BED theorized to improve outcome Lower BED theorized to decrease toxicity BED = (nd[1 + d/(α/β)]) n is the number of radiation fractions; d is the fraction size

Purpose To determine if increasing BED > 180 Gy with HDR-BT is correlated with outcomes or late toxicities

Methods Meta-analysis that includes 12,311 prostate cancer patients (n) from 81 studies (N) published from 1970 to 2012 PICOS / PRISMA selection protocol

non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and/or boost Methods Patients Men with localized (T1-T2, N0-Nx, M0) and locally advanced (T3-T4, N0-Nx, M0) prostate cancer Intervention non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and/or boost Control Either no control group (i.e. intervention as a monotherapy); or a multi-arm study that contains the intervention Outcomes Actuarial FFBF @ 5-years, stratified by risk groups Late RTOG toxicities, GI and GU Study design Large (n > 150), prospective Small, retrospective studies included to account for variability in fractionation schedules and BEDs

non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and/or boost Methods non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and/or boost Studies identified through database searching (N = 201) Studies identified through database searching (N = 270) Identification Studies after duplicates and updates removed (N = 190) Studies after duplicates and updates removed (N = 255) Studies screened (N = 190) Studies screened (N = 255) Screening Studies excluded (N = 130) Studies excluded (N = 123) Full-text studies assessed for eligibility (N = 60) Studies assessed for eligibility (N = 102) Updates of HFRT trials from national meetings (N = 1) Full-text studies excluded, with reasons retrospective, not a clinical trial, or non-randomized (31) focus on stereotactic body radiation therapy or high dose rate brachytherapy (9) report on dosimetry only (8) report on efficacy or safety only (1) Full-text studies excluded, with reasons focus on stereotactic body radiation therapy (6) report on dosimetry, physics only (12) report on acute outcomes / toxicity only (14) report not on primary or definitive HDR-BT (15) Eligibility Major trials comparing non-DE-CFRT to DE-CFRT (N = 6) Studies included in qualitative synthesis HFRT vs. non-DE-CFRT or DE-CFRT (N =12) Major trials comparing non-DE-CFRT to DE-CFRT (N = 6) Studies included in qualitative synthesis HDR-BT mono (N = 11 [not exclusive]) HDR-BT boost (N = 44 [not exclusive]) Included n = 4,028 n = 8,283

non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and/or boost Methods non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and/or boost (n) (d) [1 + ] (d) (α/β) CALCULATE BEDs Mixed effects regression models used to estimate weighted linear relationships between BED and: Outcomes (i.e. 5-year FFBF) Late toxicity

Outcomes 5-year FFBF vs. BED at α/β of 1.5 low-risk intermediate-risk high-risk

Low-risk non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and boost 5-year FFBF (%) BED with α/β of 1.5 (Gy) non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and boost

Intermediate-risk non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and boost 5-year FFBF (%) BED with α/β of 1.5 (Gy) non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and boost

High-risk non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and boost 5-year FFBF (%) BED with α/β of 1.5 (Gy) non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and boost

Toxicity Late RTOG toxicity vs. BED at α/β of 3.0 Sites: Grades: Gastrointestinal (GI) Genitourinary (GU) Grades: 0: none 1: mild 2: moderate 3: severe 4: extreme

Toxicity non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and boost incidence (%) BED with α/β of 3.0 (Gy) non-DE-CFRT; DE-CFRT; HFRT HDR-BT mono and boost

Results in perspective Outcomes and toxicities of increasing BED for prostate cancer radiotherapy Study type N (n) BED (Gy) ranges at α/β Risk group 5-year FFBFs at BED range with α/β of 1.5 Late RTOG grade 3-4 toxicity at BED range with α/β of 3.0 1.5 3.0 % range Slope GU % range GI % range non-DE-CFRT; DE-CFRT; HFRT 15 (4,028) 140-200 98-133 L 75-100 † 0-7 0.61 0-13 1.54* I 60-100 4.94** H 55-85 HDR-BT monotherapy or boost 81 (8,283) 180-310 107-188 90-100 - 0.14 0-12 0.57 0-4 0.10 85-100 - 0.01 50-85 0.20 Note: * denotes p-value < 0.05 ** denotes p-value < 0.01 † denotes insufficient data reported

Results in perspective CFRT / HFRT 5-year FFBF (%) non-dose escalated dose-escalated risk group HDR-BT 90 80 70 100 60 low intermediate high 20 toxicity insert incidence (%) 100 BED at α/β of 5 150 100 140 200 300 BED at α/β of 1.5

Conclusions: tumor perspective Increasing BED (at α/β of 1.5) from 140 to 200 Gy is associated with improvements in percent FFBF, specifically in intermediate-risk patients Further increments of BED above 200 Gy are not significantly associated with improved FFBF among any risk groups

Conclusions: normal tissue perspective Increasing BED (at α/β of 3.0) from 98 to 133 Gy is associated with increased toxicity Increasing from 107 to 188 Gy is not associated with a change in toxicity

END nicholaszaorsky@gmail.com