ESTRO 2025 Congress Report
I. Session on transformative combined modality treatment strategies Highlights from Proffered Papers
Osimertinib after chemoradiotherapy in unresectable stage III EGFRm NSCLC: chemoradiotherapy regimens and radiation pneumonitis in LAURA (NCT0352154)
Terufumi Kato, from Yokohama, Japan, reported new data from the LAURA study regarding chemoradiotherapy (cRT) regimens and radiation pneumonitis.
As per protocol, patients ³ 18 years old (³20 years in Japan), with stage III unresectable non-small-cell lung cancer (NSCLC) that harboured mutations to the epidermal growth factor receptor (EGFR) gene, were included. Definitive cRT with a radiation dose of 60Gy ± 10% (54-66Gy) was required; a mean lung dose (MLD) of < 20Gy and the proportion of the lung volume receiving at least 20Gy (V20) of < 35% were recommended but not required. Both concurrent (cCRT) and sequential (sCRT) chemoradiotherapy were used.
Patients who had not progressed after definitive cRT were randomised 2:1 (within six weeks of the end of radiation treatment) to receive either 80mg/day of osimertinib or placebo, until blinded independent central reviews confirmed progression or discontinuation. Patients with asymptomatic radiation pneumonitis (RP) were included.
A total of 216 patients were randomised (143 to osimertinib, 73 to placebo). Baseline characteristics were generally balanced between the osimertinib/placebo arms: cCRT 92/85%, sCRT 8/15%, respectively. All patients received the required radiation dose (60Gy ± 10%): 60Gy at 2Gy/fx was the most common regimen (55% each arm). For the dosimetry data, in the osimertinib and placebo arms, the MLD was 15/15Gy, lung V20 26/26% and lung V5 54/52%, respectively.
Both cRT and sCRT demonstrated the benefit of osimertinib vs. placebo. Overall response rate was higher in the osimertinib arm (57%) than in the placebo arm (33%). RP was the only radiation-related adverse event reported (48% osimertinib arm, 38% placebo): no G4/G5 level events were reported, G1/G2 were the most common. Two patients had G3 RP in the osimertinib arm, zero in the placebo arm. RP caused a few discontinuations: 5% in the osimertinib arm vs. 3% in the placebo arm.
In the univariate analysis, factors associated with RP (all grades) were: time from end of cRT to randomisation, lung V5, lung V20, MLD, radiotherapy technique used (intensity-modulated/volumetric modulated arc vs. 3D) and response to prior cRT.
In the multivariate analysis, lung V5, radiotherapy technique, response to prior cRT and stage of the disease were associated with RP (all grades).
The mandated radiation dose was met in all patients and was balanced in the treatment arms. The RP was of low grade and manageable, with low rates of discontinuation.
Exploratory multivariate analyses found that patients with a lung V5 of £ 54%(vs. ³54%), treatment via intensity-modulated radiotherapy (vs. 3D) and stable disease/Non Evaluable response (vs. complete or partial response) to prior cRT were more likely to experience RP (all grades).
II. Highlights from “LUNG 3” Session: poster discussion
Impact of prescription to the PTV on GTV dose heterogeneity in lung SBRT: insight from EORTC studies
Treatment of lungs by Stereotactic body radiation therapy (SBRT) requires dose prescription to the planning target volume (PTV) according to the International Commission on Radiation Units and Measurements recommendation. However, this may cause significant variability in the dose delivered to the gross tumour volume (GTV) because the tissue is non-homogeneous and moves; yet, accurate GTV dosages must be delivered for lung SBRT.
The European Organisation for Research and Treatment of Cancer (EORTC) investigated the dose distribution vs. heterogeneity at GTVs when prescribing dose at PTVs for lung SBRT. Hertsyk Volha, clinical research fellow at the radiotherapy quality assurance unit of the medical department at EORTC, presented the results.
Data had been extracted from four EORTC studies (1882 OligoCare, 1702HALT, 1945OligoRare, 22113 LungTech). A total of 232 lung lesions were identified in 167 patients who were involved in the studies listed above. PTV and mean GTV doses, dose calculation algorithms and target locations were analysed. Since the doses prescribed for PTVs varied among the trials, the relative dose difference was used to make comparative results.
The relative dose difference between the PTV prescribed dose and mean GTV dose was highest in both the 1882 cohort and in the OligoCare study (with median differences of 29.1% and 21.8% respectively), while for the LungTech trial there was a lower median difference (19.5%) and the LungTech study had the smallest median difference (13.7%).
These results suggested that direct prescription of the dose to the GTV while optimising the plan to the PTV could provide more consistent reports of the doses delivered within tumours, reduce dose discrepancies and enhance treatment precision by mitigating the influence of surrounding low-density tissue on final dose evaluation.
Data for the clinical impact of this strategy are awaited; analyses are ongoing.
Effect of SABR on detection of circulating tumour DNA in early-stage lung cancer: interim results from a prospective, multi-centre trial
Stereotactic ablative radiotherapy (SABR) or SBRT is a radical option for the treatment of patients with early-stage lung cancer who are ineligible for surgery.
Most of these patients have severe comorbidities and cannot undergo biopsies, so they are treated only with radiological and clinical signs of cancer without histological confirmation.
The use of blood-based liquid biopsies of circulating DNA (ctDNA) in patients without tissue diagnosis might allow pathological confirmation of cancer and further molecular testing. However, usually the rates of detection of ctDNA are low in early-stage NSCLC.
Young Sympascho, from the Verspeteen Family Cancer Centre in Canada, presented the results of an interim analysis of a prospective, multi-centre trial of the effect of SABR on ctDNA, with the hypothesis that liquid biopsies might show higher ctDNA detection rates once SABR had been started.
In the multi-institutional study, patients who were due to receive SABR for early-stage lung cancer were enrolled into two cohorts:
- patients with suspected stage IA/IIA NSCLC and a pre-radiotherapy likelihood of malignancy of ³60% according to the Herder and/or Brock models (n=45); or
- patients with biopsy-proven NSCLC (n=30-60).
The ctDNA was collected prior to the first fraction of SABR and 24-72 hours after (range 7.5-18Gy); ctDNA was detected through the use of the SHIELDING ULTRA MRD panel for agnostic tumours.
The primary endpoint of the study was to evaluate whether minimal residual disease could be used for prognosis and predictor of early recurrence in SABR patients.
In the interim analyses, the secondary objective, which was to assess the impact of SABR on ctDNA detection rates, was reported: plasma samples for ctDNA before and after the first SABR fractions were tested for 56 patients. After quality control, 53 paired samples were analysable.
Of the 56 patients, 33 (59%) were male and 21 (41%) female. A total of 24 (43%) patients had diagnoses that had been confirmed from tissue: 14 (58%) showed non-squamous histology, while 10 (42%) were squamous.
Pre-SABR ctDNA detection rates were 32% compared with post-SABR rates of 43%; 47% were detected either pre- or post-SABR. For squamous cell carcinomas, the pre-SABR rates were 44% (4/9) and the post-SABR rates were 56% (5/13). For adenocarcinomas, the pre-SABR rates were 15% (2/13) and the post-SABR rates were 39% (5/13).
The results showed that the rate of ctDNA detection increased when both the pre- and post-SABR samples were tested; the post-SABR (within 24-72h) may have higher detection rates.
Blood samples to test for ctDNA are planned from three months after completion of SABR and then every three months up to 24 months after SABR; accrual is ongoing.
Valeria Dionisi
Radiation oncologist
Radiation Oncology Department
University and Hospital Trust
Verona, Italy
valeria.dionisi@gmail.com
valeria.dionisi@aovr.veneto.it