We use cookies necessary to giving you a better online experience. By using our website you consent to all cookies in accordance with our Privacy Policy.
Yes, I accept
Menu
Search Opener
Vienna, Austria

ESTRO 2023

Session Item

Radiomics, modelling and statistical methods
Poster (Digital)
Physics
Can Radiomics support Early Regression Index in predicting rectal cancer response to MRgRT?
Luca Boldrini, Italy
PO-2090

Abstract

Can Radiomics support Early Regression Index in predicting rectal cancer response to MRgRT?
Authors:

Luca Boldrini1, Giuditta Chiloiro2, Davide Cusumano3, Poonam Yadav4, Gao Yu5, Angela Romano2, Antonio Piras6, Lorenzo Placidi7, Sara Broggi8, Francesco Catucci3, Jacopo Lenkowicz7, Luca Indovina7, Michael F Bassetti9, Yingli Yang10, Claudio Fiorino8, Vincenzo Valentini2, Maria Antonietta Gambacorta2

1Fondazione Policlinico Universitario "A. Gemelli" IRCCS , Radiology, Radiation Oncology and Hematology, Rome, Italy; 2Fondazione Policlinico Universitario “A. Gemelli” IRCCS, Radiology, Radiation Oncology and Hematology, Rome, Italy; 3Mater Olbia Hospital, Radiotherapy, Olbia, Italy; 4Northwestern Memorial Hospital, Radiation Oncology, Chicago, USA; 5University of California Los Angeles, Radiological Sciences, Los Angeles, USA; 6Villa Santa Teresa, UO Radioterapia Oncologica, Bagheria, Italy; 7Fondazione Policlinico Universitario “A.Gemelli” IRCCS, Radiology, Radiation Oncology and Hematology, Rome, Italy; 8San Raffaele Scientific Institute, Medical Physics, Milan, Italy; 9University of Wisconsin , Department of Human Oncology School of Medicine and Public Heath, Wisconsin , USA; 10University of California Los Angeles, Radiological Sciences, Los Angeles, USA

Show Affiliations
Purpose or Objective

Early Regression Index (ERITCP) is able to efficaciously predict pathological complete response (pCR) on magnetic resonance (MR) images of patients with Locally Advanced Rectal Cancer (LARC). This study aims to investigate the potential role of Radiomics in adding discriminative power to ERITCP in patients undergoing neoadjuvant chemoradiotherapy on a 0.35 T MR Linac.



Material and Methods

LARC patients (cT2-4 and/or cN0-2, cM0) treated with MR-guided Radiotherapy (MRgRT) with different schedules were retrospectively enrolled from three international institutions and divided in training and validation set for a pCR predictive model.

Biologically effective dose (BED) conversion was used to compensate for the different RT schemes. 0.35T T2*/T1-weighted MR images were acquired during simulation and prior to each treatment fraction.

The rectal Gross Tumor Volume (GTV) was manually contoured on the 0.35 T MR images corresponding to the BED levels of 0,13,26,40,53 and 59 Gy and radiomics analysis was performed using MODDICOM, a dedicated image analysis platform compliant with the IBSI recommendations.

Multiple logistic regression models, combining ERITCP with radiomic features, were calculated and compared with the model obtained using ERITCP as unique variable. Predictive performances were evaluated using Receiver Operating Characteristic (ROC) curve and model comparison was performed using Delong test. Results were evaluated on a separated validation set.

Results

A total of 91 patients were enrolled. Patients treated from April 2015 to February 2020 were considered as training set (58), while patients treated from March 2020 to December 2021 were considered for validation set (33).

Overall, pCR (defined as absence of viable tumour cells in both primary tumour and nodal pathological specimens, i.e. pT0N0) was observed in 25 cases (27.4%).

A total of 990 quantitative image features were extracted and processed.

A total of 118 nested models were then calculated, combining each significant radiomic feature with ERITCP. The model showing the highest performance was obtained combining ERITCP with the 10th percentile of the grey-levels histogram, calculated on the GTV of the MR image acquired when a BED value of 40 Gy was reached.

For the training set, the resulting area under curve (AUC) values of the ERITCP model and the combined model were 0.94 (95%CI: 0.88-0.99) and 0.98 (95%CI: 0.95-1) (p=0.04) respectively, while 0.89 and 0.92 were obtained on the validation set (Fig.1).


Conclusion

The integration of radiomic features with ERITCP is able to improve pCR prediction in LARC patients. The combined model can be particularly useful in borderline situations where ERITCP is close to the cut-off threshold.