Optimising oligometastatic disease management with circulating tumour DNA: a correlative analysis from the phase II randomised EXTEND trial

ESTRO 2026 Congress report

A translational analysis evaluating circulating tumour DNA (ctDNA) as a prognostic and treatment-stratifying biomarker for oligometastatic disease was presented at the ESTRO 2026 annual meeting, drawing on plasma samples from the randomised phase II EXTEND trial.

Alexander Sherry from the United States presented a correlative analysis from EXTEND, the phase II basket trial that randomised 350 patients with one to five metastases across six tumour categories (pancreas, breast, kidney, prostate on continuous androgen deprivation therapy [cADT], prostate on intermittent androgen deprivation therapy [iADT], and an "Other" basket) to metastasis-directed therapy (MDT) plus standard of care (SOC) versus SOC alone. Current selection of patients for MDT relies on counting radiologic lesions, an approach that may misclassify true micrometastatic disease burden. The investigators hypothesised that ctDNA detection would stratify prognosis more accurately and identify patients more likely to benefit from MDT.

Plasma was collected at enrollment, at the 3-month follow-up, and at progression, and analysed with the tumour-agnostic, methylation-based Guardant Reveal assay. ctDNA was detected at enrollment in 115 of 237 patients (49%), most often in pancreatic cancer (70%) and least often in prostate cancer (35%). Cox models were adjusted for treatment arm, basket, and stratification factors.

ctDNA positivity at enrollment was strongly associated with worse outcomes across all baskets, with no evidence of heterogeneity: progression-free survival (PFS) HR 2.51 (95% CI, 1.73-3.66; p<0.001) and overall survival (OS) HR 2.05 (95% CI, 1.20-3.67; p=0.009). MDT improved PFS regardless of ctDNA status (ctDNA-positive HR 0.61 [0.38-0.98], p=0.04; ctDNA-negative HR 0.41 [0.22-0.76], p=0.005). At the 3-month landmark in 188 patients with paired plasma, ctDNA positivity remained prognostic for PFS (HR 1.94 [1.25-2.99], p=0.003) and OS (HR 2.59 [1.46-4.72], p=0.001). Patients whose ctDNA cleared between enrollment and the 3-month timepoint had substantially better OS than those with persistent positivity (HR 0.25 [0.08-0.64], p=0.003).

This is the largest prospective evaluation of ctDNA in a randomised trial of MDT for oligometastatic disease. The take-home is twofold. ctDNA detection consistently and independently identifies patients at higher risk of progression and death across tumour types, supporting its use to refine the operational definition of oligometastasis beyond lesion counting. ctDNA clearance, in turn, may serve as an early surrogate endpoint for OS, with practical implications for trial design. The data do not, however, support withholding MDT from ctDNA-positive patients: benefit was preserved in both subgroups. Prospective validation trials are being planned.

The primary results of the EXTEND trial were published in parallel in the Journal of Clinical Oncology (https://doi.org/10.1200/JCO-25-02856).

 

Early Toxicity of elective nodal RT for Oligorecurrent Pelvic/Para-aortic nodes in Prostate Cancer: a randomised phase 3 trial OLIGOPELVIS-2/GETUG P12.

A randomised phase III trial comparing salvage elective nodal radiotherapy plus intermittent androgen deprivation therapy (ADT) with intermittent ADT alone for oligorecurrent pelvic and para-aortic prostate cancer was presented at the ESTRO 2026 annual meeting.

Stéphane Supiot from France presented the 18-month toxicity results of OLIGOPELVIS-2 / GETUG P12, a multicenter randomised phase III trial conducted across 17 French centres. Eligible patients had histologically proven prostate adenocarcinoma with recurrence after radical local treatment (surgery and/or radiotherapy), no more than five pelvic or para-aortic lymph node recurrences on choline or prostate-specific membrane antigen (PSMA) PET imaging, performance status 0-1, and testosterone above 6 nmol/L. The initial upper field limit was the aortic bifurcation; following an analysis of relapse patterns from the OLIGOPELVIS GETUG P07 study showing frequent para-aortic relapses, a December 2021 amendment extended coverage to the renal arteries. Patients were randomised 1:1 to six months of intermittent ADT (iADT) alone (arm A) or to six months of iADT plus salvage elective nodal radiotherapy (ENRT, arm B). ENRT delivered 54 Gy in 30 fractions of 1.8 Gy to the pelvic and para-aortic nodes, with a simultaneous integrated boost to 66 Gy at 2.2 Gy per fraction to PET-positive pathological nodes, starting after three months of iADT. Patients with prior pelvic radiotherapy were eligible only if the cumulative dose to the first target node remained below 20 Gy.

Between December 2018 and May 2023, 256 patients were enrolled. Eight patients randomised to arm B did not receive ENRT and were excluded from the safety population, leaving 127 patients in arm A and 121 in arm B. In arm B, prior prostatic bed radiotherapy after prostatectomy had been delivered to 45% and, prior primary prostate radiotherapy to 10%, and 39 patients (32%) received para-aortic irradiation.

At 6 months, grade ≥2 gastrointestinal (GI) toxicity was 0% in arm A versus 19.8% in arm B (p<0.0001), and grade ≥2 genitourinary (GU) toxicity was 2.4% versus 9.9% (p=0.02). By 18 months, grade ≥2 GI toxicity had largely resolved (0.8% vs 4.1%, not significant), while grade ≥2 GU toxicity remained modestly higher in arm B (4.1% vs 10.7%, p=0.04). Grade ≥2 urinary incontinence at 18 months was 8.3% with ENRT versus 2.5% without (p=0.046). No grade 4 events and no treatment-related deaths were observed. Asthenia and erectile dysfunction did not differ significantly between arms. Subgroup analysis showed no excess grade ≥2 toxicity in previously irradiated patients, although the dose junction between prior prostate or prostate-bed radiotherapy and the new pelvic fields warrants vigilance. Among the 39 patients receiving para-aortic ENRT, the only signal was a transient increase in grade ≥1 upper-GI symptoms at 6 months (25.6% vs 9.8%, p=0.02) that resolved by 18 months.

This is the first randomised phase III trial of ENRT with a simultaneous boost to pathological nodes in nodal oligometastatic prostate cancer recurrence, and the first to prospectively report para-aortic irradiation in this setting. Toxicity was acceptable, slightly higher than in PEACE V-STORM (which used 45 Gy to the pelvis), but tolerable across all subgroups, including previously irradiated patients and those receiving para-aortic coverage. Longer follow-up and the trial’s primary endpoint, biochemical and clinical failure, are awaited.

The full early-toxicity results are published in Radiotherapy and Oncology (https://doi.org/10.1016/j.radonc.2026.111533).

 

Oded Icht, MD, MBA
Staff Radiation Oncologist
Davidoff Cancer Institute, Rabin Medical Centre, Petach-Tikva, Israel
Gray Faculty of Medical and Health Sciences, Tel-Aviv University, Israel
odedicht@gmail.com
www.linkedin.com/in/oded-icht-md-mba-ba4a97269