Programming for Off-Line Adaptive Radiotherapy

Dattoli Cancer Center

August 2, 2022

The main difference between real-time and off-line image-guided adaptive radiotherapy is that the latter relies on software to control radiation therapy. For example, real-time image-guided adaptive radiotherapy will require a more sophisticated model that will have stopping rules to pause treatment when the deviation from the treatment plan reaches a certain threshold. Aside from the benefits of real-time image-guided adaptive radiotherapy, the latter will also require a more complex model that can perform more advanced tasks, such as patient monitoring.

Adaptive radiotherapy

Off-line adaptive radiotherapy is more complex than its online counterpart. In the past, programming for offline adaptive radiotherapy relied mainly on manual calculations, but now it requires the assistance of sophisticated software. For example, MIM software allows clinicians to project final doses and tabulate daily doses. This helps them to design boost plans and make off-line adaptive radiotherapy clinically manageable. This article will examine the advantages and disadvantages of off-line adaptive radiotherapy.

CBCT is the most common imaging modality used in oncology. It helps radiotherapists tailor a radiotherapy plan to the patient’s body mass. The amount of soft tissue surrounding the target can vary by several centimeters, so using the data from CBCT scans is an effective method. CBCT images are taken on a daily basis to create a library of plans for an individual patient.

The goal of intensity-modulated radiation therapy (IMRT) is to deliver high doses to tumors while sparing healthy tissue. Optimization-based treatment planning produces sharp dose gradients between tumors and healthy tissues. Furthermore, the random shifts during the treatment process can cause large dose differences between the two groups. As a result, IMRT treatment plans deliver the dose in small fractions over a period of 35 days.

Model formulation

The model formulation for off-line adaptive radiotherapy requires careful consideration of the various aspects of the treatment. Generally, it is recommended to choose the online adaptation well before the initiation of treatment. It also requires careful pre-planning and physician and physics participation. In most cases, this technique is best applied in predictable situations, such as stereotactic body radiotherapy. This technique is particularly useful when daily anatomic changes are predictable, such as those associated with bowel filling or peristalsis.

Off-line ART relies on high-quality imaging to detect changes that require offline treatment. Some changes are apparent without images, such as weight loss or volume changes in superficial tumors. Other changes, such as internal anatomy, require imaging for visualization. Standard cone-beam CT technology, which is available in most modern medical linear accelerators, is useful for detecting changes in the volume of a lung tumor, an exophytic lesion, or a fiducial marker. In addition, planar X-ray imaging can detect alterations in fiducial markers. And, with auxiliary detector systems, planar X-ray imaging can detect fiducial marker motion.

The study also compared adaptive plans to non-adaptive ones. The results demonstrated that the adaptive plan reduced the amount of time patients spent on the treatment table and improved resource allocation. The study’s clinical threshold for adaptation will depend on the site of treatment, type of treatment, and organ-at-risk. Pre-specified criteria can also prevent indecision at the time of delivery. They are an important step in optimizing off-line adaptive radiotherapy.

Constrained optimization methods

Off-line adaptive radiotherapy is a form of radiation therapy in which the treatment plan is chosen by a radiation oncologist after considering various trade-offs. It is administered over multiple sessions and involves a substantial fraction of cancer patients. Optimization models help physicians achieve optimal outcomes for each patient. For example, an optimization model can optimize the dose for a particular tumor while keeping the healthy tissues protected.

Optimization methods that use the LQR model to determine treatment plans are known as stochastic control. The authors estimate the tumor response with either a log-linear cell kill model or a standard LQ model. Their goal is to minimize the number of tumor cells at the end of the treatment. Their method focuses on beam intensities and fixed sessions to achieve optimal treatment planning. Constrained optimization methods for off-line adaptive radiotherapy are not only based on tumor response criteria but are also computationally efficient.

The planning process for offline ART is not much different from the standard clinical workflow. Typically, the angle of the beam in an adaptive plan is close to the patient’s initial position during simulation. However, since the patient’s organs and targets will be moving around during treatment, a robust plan is required to account for these changes. For instance, gastrointestinal OARs can move closer to the target than their original position on a given day. Non-adaptive planning places a higher priority on the OARs that are closest to the target, but this is not the case with online adaptation.