Early radiologic and metabolic tumour response assessment during combined chemo-radiotherapy for locally advanced NSCLC

Marie Tvilum; Marianne Marquard Knap; Lone Hoffmann; Azza Ahmed Khalil; Ane L Appelt; Ate Haraldsen; Markus Alber; Cai Grau; Hjørdis Hjalting Schmidt; Maria Kandi; Marianne Ingerslev Holt; Christina Maria Lutz; Ditte Sloth Møller

doi:10.1016/j.ctro.2024.100737

Early radiologic and metabolic tumour response assessment during combined chemo-radiotherapy for locally advanced NSCLC

Clin Transl Radiat Oncol. 2024 Jan 23:45:100737. doi: 10.1016/j.ctro.2024.100737. eCollection 2024 Mar.

Authors

Marie Tvilum^{1

2}, Marianne Marquard Knap¹, Lone Hoffmann^{1

3}, Azza Ahmed Khalil¹, Ane L Appelt^{4

5}, Ate Haraldsen⁶, Markus Alber^{7

8}, Cai Grau², Hjørdis Hjalting Schmidt¹, Maria Kandi¹, Marianne Ingerslev Holt⁹, Christina Maria Lutz¹, Ditte Sloth Møller^{1

3}

Affiliations

¹ Department of Oncology, Aarhus University Hospital, Denmark.
² Danish Center for Particle Therapy, Aarhus University Hospital, Denmark.
³ Department of Clinical Medicine, Faculty of Health Sciences, Aarhus University, Denmark.
⁴ Leeds Institute of Medical Research at St James's, University of Leeds, United Kingdom.
⁵ Leeds Cancer Centre, Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom.
⁶ Department of Nuclear Medicine and PET-centre, Aarhus University Hospital, Denmark.
⁷ Department of Radiation Oncology, Heidelberg University Hospital, Germany.
⁸ Heidelberg Institute for Radiation Oncology (HIRO), Heidelberg University Hospital, Germany.
⁹ Department of Clinical Genetics, Sygehus Lillebaelt, Vejle, Denmark.

Abstract

Background: The role of early treatment response for patients with locally advanced non-small cell lung cancer (LA-NSCLC) treated with concurrent chemo-radiotherapy (cCRT) is unclear. The study aims to investigate the predictive value of response to induction chemotherapy (iCX) and the correlation with pattern of failure (PoF).

Materials and methods: Patients with LA-NSCLC treated with cCRT were included for analyses (n = 276). Target delineations were registered from radiotherapy planning PET/CT to diagnostic PET/CT, in between which patients received iCX. Volume, sphericity, and SUVpeak were extracted from each scan. First site of failure was categorised as loco-regional (LR), distant (DM), or simultaneous LR+M (LR+M). Fine and Gray models for PoF were performed: a baseline model (including performance status (PS), stage, and histology), an image model for squamous cell carcinoma (SCC), and an image model for non-SCC. Parameters included PS, volume (VOL) of tumour, VOL of lymph nodes, ΔVOL, sphericity, SUVpeak, ΔSUVpeak, and oligometastatic disease.

Results: Median follow-up was 7.6 years. SCC had higher sub-distribution hazard ratio (sHR) for LRF (sHR = 2.771 [1.577:4.87], p < 0.01) and decreased sHR for DM (sHR = 0.247 [0.125:0.485], p < 0.01). For both image models, high diagnostic SUVpeak increased risk of LRF (sHR = 1.059 [1.05:1.106], p < 0.01 for SCC, sHR = 1.12 [1.03:1.21], p < 0.01 for non-SCC). Patients with SCC and less decrease in VOL had higher sHR for DM (sHR = 1.025[1.001:1.048] pr. % increase, p = 0.038).

Conclusion: Poor response in disease volume was correlated with higher sHR of DM for SCC, no other clear correlation of response and PoF was observed. Histology significantly correlated with PoF with SCC prone to LRF and non-SCC prone to DM as first site of failure. High SUVpeak at diagnosis increased the risk of LRF for both histologies.

Keywords: Early treatment response; Non-small cell lung cancer; Pattern of failure.