Radiation Oncologist | MSc Physics | PhD AI & Quantum Computing / Oncologic Imaging | Views are my own
Québec, Canada
Joined January 2012
- Tweets1.7K
- Following663
- Followers507
- Likes3.3K
Pinned Tweet
Happy to share our recent publication on the use of quantum technology for machine learning to differentiate between brain metastases and high-grade gliomas on MRI. 1/3
doi.org/10.1007/s10278-023-0…
Tony Felefly retweeted
⚡️ Phase 2 trial in RCC with inferior vena cava tumor thrombus: neoadjuvant stereotactic ablative radiotherapy (SAbR) followed by surgery (n=23).
1-year recurrence-free survival: 71.4% — above the 55.7% historical benchmark. Thrombus downstaging in 22%.
A promising approach for one of the most challenging complications in kidney cancer.
#KidneyCancer #RCC @JAMAOnc
jamanetwork.com/journals/jam…
Worth reading: a sharp analysis of PROTEUS trial results by @DrSpratticus 👇
nature.com/articles/s41571-0…
Tony Felefly retweeted
Nice 📄 from PRIAS: MRI🧲 matters for AS selection. Repeat biopsies mostly need to focus on the MRI lesion.
“Progression” 🟰correcting initial undersampling?
🤔 How good is targeting in targeted biopsy?
@dr_coops @mleapman @NikiSushentsev
sciencedirect.com/science/ar…
Tony Felefly retweeted
An observational paper reports an MLE and 95% CI for the effect β. What it estimated is θ = β + b: effect plus bias. The MLE targets θ and the CI covers θ. The default format has no slot for b, so it quietly sets b = 0. 🧵
Tony Felefly retweeted
Frequentist inference has no answer for β here: only θ = β + b is identified, so the P-value tests θ = 0, not β = 0. Bayes puts a prior on b
Not sure I understand how underlying physics make sense here.
With EBRT, the Compton effect largely predominates, with any photoelectric effect contribution being extremely low. Even backscatter is governed by the Compton effect.
I agree this might be relevant for LDR brachy, but not sure it can explain things for SBRT. Also, modern Type-C calculation algorithms model backscatter already, and of course tissue density.
Am I missing something?
The paper is however interesting. Results are hypothesis-generating, but many confounders exist here, such as Pelvic ENI imbalance, and a higher moderate dose to the bladder.
@jryckman3 @5_utr @toddscarbrough @SprakerMDPhD
Tony Felefly retweeted
The absence of (positive) evidence in this study may not indicate evidence of absence re: benefit for the approach across breast OMD situations. While perhaps not standard, it may still be very reasonable to consider case by case as there is a lot of nuance with radiotherapy (not all #SBRT is the same re: dosing and planning) as well as trial design and the biological heterogeneity of both biomarker profiles and oligometastatic presentations (volume, number of mets, number of organ systems, temporality) that merit consideration before jettisoning the whole approach. Lots of considerations discussed in thread below and a #radonc perspective on the evidence in the breast OMD space will be discussed at an upcoming virtual event (sabrcanada.org)
@jryckman3
👏🎗️ Congrats to @Docace911 @NRGonc on BR002 ascopubs.org/doi/full/10.120…
⭐️Prelim thoughts
- primary must be controlled
- brain mets excluded
- only 22% de novo
- 1/3 pts progressed in RT field❓
One attempts to contrast these w/ EXTEND&OLIOGMA. Regardless, very diff populations.
Look at those beautiful "Focal Radiotherapy" volumes!
And people keep saying the Radiation Oncology community is against focal therapy!
Sigh!
Just prove it's equivalent to whole gland therapy, and we're all the way in!
#focaltherapy
2/6 | Patient Selection and Tx
N=62 participants from 2017-2023 with IPR after definitive RT from 3 institutions (Loyola, UVA, and UTSW).
• 60% had EBRT & 24% had LDR
• Median duration from prior RT was 8 years
HDR: Goal of 30 Gy in 1-2 fractions, limiting dose to prioritize normal tissue structures. We used a focal approach for planning - utilizing information from mpMRI, PET/CT and biopsy findings.
This was primarily a ReRT monotherapy approach, as only 8% received concurrent ADT.
Tony Felefly retweeted
1/6:
I am excited to share our multi-institutional F-SHARP phase I/II trial, now published in @JCO_ASCO!
doi.org/10.1200/JCO-25-02730
There are limited prospective data defining the optimal management of intraprostatic radiorecurrence (IPR) after definitive RT for prostate cancer.
We asked:
1. Is dose-escalated focal salvage high dose rate brachytherapy safe and effective for IPR?
2. Can Decipher (@Veracyte) identify patients with higher risk of recurrence with salvage ReRT that may benefit from therapeutic intensification?
🧵👇
Tony Felefly retweeted
CURB also reported no progression in any irradiated lesion across its SBRT arm, including 24 breast cancer pts. Shorter follow-up and a more heavily treated oligoprogressive population, but this adds to my question about BR002’s 20/60 pts progressing within the ablated area. Are we counting different events, or seeing a real difference in local control?
I’m not suggesting this would have made BR002 positive.
I’d like to understand what the 20 events represent: regrowth of treated targets, new lesions within the treated area, or something else? And how does that relate to the 6.7% inside-area first-failure figure reported in 2022?
Just trying to understand
Tony Felefly retweeted
New in the @IJROBP : breast RT + SIB in 5 fractions, with low objective late side effects at 5 years.
Congratulations to Maia Dzhugashvili this paper forms part of her PhD thesis and all coauthors!
🔗 authors.elsevier.com/sd/arti…
Tony Felefly retweeted
Drugs can barely penetrate GBMs. We built a nanomotor that crosses the BBB, senses the tumor, and swims toward it, powered by hydrogen peroxide, which GBM produces in excess.
3x deeper penetration. 4.3x more potent.
Out now in @MacroBioscience
onlinelibrary.wiley.com/doi/…
Tony Felefly retweeted
i-STOP randomized 60 men with PSMA PET-only oligometastatic prostate cancer: 3-year biochemical progression was 21% with SBRT plus intermittent ADT vs 60% with iADT alone. Feasibility was the primary endpoint, so larger efficacy trials are needed.
amportal.astro.org/sessions/…
Tony Felefly retweeted
Updated LUNAR analysis out in @EUplatinum. Now with a median follow-up of 32 months, exploratory analyses confirm the addition of 177Lu-PNT2002 to SBRT improves polymetastatic PFS (6 or more new lesion), castrate disease-free survival, and CT-based PFS (i.e., conventional MFS) authors.elsevier.com/sd/arti…
Tony Felefly retweeted
With k=2, CIs overlap b/c they’re noisy draws of the same mechanism. Model joint trial effects across histologies as MVN. Partial pooling shrinks OLIGOMA’s winner-curse & BR002’s noise toward a shared μ.
Tony Felefly retweeted
🚨 PSMA-guided RT intensification for #ProstateCancer 🚨
@EuropeanUrology
👥 262 randomized (253 treated), 4 strata: salvage post-RP, high-risk, oligomet, salvage post-RT
📊 Phase 2 RCT, SOC RT vs PSMA-PET-guided RT to all detected sites
⏱️ Median f/u 66 mo
⚠️ Primary 5-yr failure-free survival 66% vs 62%, HR 0.80, p=0.14 (not met)
✅ BCFS favored PSMAgRT: 72% vs 65% (sHR 0.70, p=0.048)
✅ High-risk 81% vs 66%; post-RP 80% vs 68% (sHR 0.54, p=0.037)
🎯 Seeing more w/ PSMA doesn't mean treating more helps everyone.
-High-risk & post-RP pts look like the target group
-Pending phase 3 PATRON.
🔗 doi.org/10.1016/j.eururo.202…
@AmerUrological @UroOnc @PCFnews @AmarUKishan @UrologyTimes @urotoday @SbrtSean @PCF_Science
Tony Felefly retweeted
It’s worth noting there is actually another completed trial that is the largest of the bunch from Egypt. They were all newly diagnosed and the results were firmly negative.
am25.astro.org/sessions/ss-2…
Tony Felefly retweeted
It’s also very interesting to relate to in field control rates from other breast studies (AVATAR. Extend-breast). Of course, cross trial comparisons are a sin, but solid tumor SABR, especially to ablative doses as per BR002, are associated with 95% in field local control
Interesting to compare with in-field local control on AVATAR and EXTEND-breast. Each MFU ~2y.
AVATAR (n=32): @drspdavid
SABR: 24/2 (43%). 20/1 (17%). 35/5 (12%).
In-field LC 100%
EXTEND breast cohort (n=43): @AlexSherryMD @ebludmir
SBRT (n=32): 16-18/1 (n=10), 20/1 (n=1), 27-30/3 (n=14), 50/4 (n=7).
LC for treated lesions 100%
Of course, cross-trial comparisons are a sin, but interesting nonetheless. Fantastic to have these data, and the investigators should be commended for these efforts! 👏 #OncTwitter #BCSM #BR002
Tony Felefly retweeted
Major Adverse Cardiovascular Events in Men with High-risk Localized or Metastatic Prostate Cancer Starting Long-term Hormone Therapy in the STAMPEDE Trial - European Urology europeanurology.com/article/… – at Chalakudy, India
