Radiation Therapy has evolved over the past two decades and its precision and accuracy have improved incredibly. This was possible with the introduction of various imaging techniques like contrast CT scan, multi-parametric MRI, and PET CT scans, along with computerized radiotherapy planning techniques like IMRT (Intensity Modulated Radiation Therapy), VMAT (Volume Modulated Arc Therapy), and IGRT (Image Guidance Radiation Therapy) with onboard imaging technique like CBCT (Cone Beam CT scan) which is an in-built CT scan in the Linear Accelerator.

With the above-mentioned techniques, we could only assess the position of the patient and dose contributing to the normal structures on that particular treatment session. However, Radiation Therapy is usually delivered over a six to seven weeks period and we know that the tumor and the patient changes in shape over this period. This dynamic change in position of tumor and organs at risk (OAR) is due to various factors like movement with respiration, differential filling of rectum and bladder, change in the shape of tumor (either shrinkage or progression), weight loss of the patient.

All these factors can decrease the precision of the radiation therapy and the cumulative dose delivered to a particular organ or the tumor cannot be accurately estimated, which may decrease the therapeutic ratio.

With advancements in IGRT, we can identify anatomical changes at the earliest and do re-planning on newly acquired planning CT scans. However, this process does not allow us to gain knowledge about the delivered dose, toxicity rates, and benefit of adaptation.

Adaptive RadioTherapy (ART) is a radiation process that addresses patient-specific treatment variations, including systematic changes in weight loss, tumor and organ geometry, and biologic response as well as stochastic variations such as organ deformation, filling change, and respiration or peristaltic movements.

To address the above issue, Raysearch enables us with sophisticated ART technology like Deformable Image Registration (DIR), Dose Deformation, and Dose Accumulation.  With the help of Deformable Image Registration (DIR), it is possible to do auto-segmentation (i.e delineation of target and normal structures) on the CBCT and allows assessment of daily dose. Further sophistication, allows for the accumulation of dose in the tumor and OARs throughout treatment. This empowers the radiation oncologist to decide on ART (re-planning and treatment delivery) based on accumulated dose rather than independent snapshots.

Adaptive Radiation Therapy can be binned into three major classes:

1. Offline – this is done between fractions to address systematic and progressive changes during treatment. It is mostly done in the head and neck, lung, and prostate.
2. Online – This is done immediately before the treatment fraction. It is done particularly in the abdomen and pelvis where offline ART is insufficient to account for variations
3. Real-time – which is done during the treatment of fractions. Example – MRI-guided ART in prostate cancer.

The main steps of ART are –

1. Image acquisition – ART needs adequate information of tumor, OAR delineation, accurate dose calculation which is possible with sufficient image quality of CBCT (iterative).
2. Deformable Image Registration (DIR)– This is an important step in ART to account for changes in the shape and size of internal organs between initial and adaptive planning images. The images are usually acquired at every fraction before treatment delivery to perform tasks such as deforming contours and electron density mapping. It helps in contour propagation, plan adjustment, and fractional dose accumulation. There are various qualitative and quantitative measures for evaluating the image registration accuracy to reduce the errors.
3. Dose accumulation and tracking– As the anatomy and corresponding contours change, the initial dose calculated using the planning data set may have limited accuracy and may not represent the actual dose delivered. ART provides a summary of the delivered dose, by voxel by voxel dose accumulation for each delivery time point. This “dose of the day“ is warped back to the initial planning CT scan for dose accumulation over the total fractions to date. The methodology of deforming adaptive planning images yields improved dose estimation.
4. Rapid Replanning– Replanning cases for ART involves consideration of strategy ( online vs offline), delineation of targets and OARs, time is taken for replanning, and clinical criteria for adaptation.
5. Pre-treatment planning and delivery quality assurance– Offline ART strategies follow the standard treatment plan and delivery QA approaches that EBRT planning workflow. Online ART needs automated plan checks and secondary dose calculation tools for the independent evaluation of plan quality.

Raysearch even helps with Artificial Intelligence (AI) in various components of ART, like auto-contouring and auto segmentation, registration, planning, and quality assurance. Raysearch is enabled with model-based and atlas-based segmentation which has the capability of machine learning.

We at Mahatma Gandhi Cancer Hospital & Research Centre are mainly focusing on lung, head and neck, prostate, and cervical cancers to improve the therapeutic ratio with Adaptive Radiotherapy.

A&D – represents the primary nasopharyngeal tumor and neck nodes before starting RT. B&D represents shrinkage of primary and nodes after 12 fractions. C& F represents the deformation of the image to propagate the contours and dose for the adaptive plan.

Written by

Dr. P Venkata Krishna Reddy,
Radiation Oncologist
Mahatma Gandhi Cancer Hospital and Research Institute, Visakhapatnam