Secondary Logo

Journal Logo

MR-Guided Online Adaptive Therapy

Christodouleas, John MD

doi: 10.1097/01.COT.0000516143.51689.1e
News
Free
John Christodouleas, MD

John Christodouleas, MD

A central challenge in radiation oncology is effectively targeting tumor tissue while sparing surrounding healthy tissue. For most diseases, we aim to treat with either higher total doses to improve efficacy or higher daily doses to improve treatment efficiency. We often cannot intensify treatment, however, because with external beam radiation, we know that normal tissue immediately surrounding the tumor will receive a high-dose “bath” (doses at or near the full prescription dose) and normal tissues further from the target will receive a low-dose bath (doses at a fraction of the prescription dose). Both types of dose baths are important, but the history of technical advancement in radiation oncology has focused on minimizing the volume of normal tissue that receives the high-dose bath because it is, by definition, the most biologically active.

The side effects of greatest concern and most limit the ability to treatment intensify are those strongly associated with the volume of normal tissue that receives the high-dose bath. The Table illustrates dose-limiting toxicities related to specific anatomical treatment sites, all of which are various manifestations of normal tissue death or dysfunction. Typically, radiation oncologists will not treat a patient unless the risks of such toxicities are believed to be very low. Fortunately, technical advances over the past 3 decades have enabled treatment intensification while maintaining a low risk of serious side effects by continually shrinking the volume of normal tissue subjected to the high-dose bath (Fig. 1).

This is the result of two important technical innovations. First, improved imaging technologies incorporated into the radiation therapy workflow allowed for more precise definition of the targeted tissue. To ensure the entire tumor is dosed effectively, the treatment area has historically included a margin of normal tissue around the tumor to account for uncertainty about its location and shape when the beam is on. A major leap forward in radiation therapy occurred with the introduction of CT scans for treatment planning, enabling us to shift from 2D simulations (which use bone landmarks to estimate tumor locations) to far more accurate 3D anatomical models. More recently, high-quality 2D imaging and even forms of CT scanners are being used at the point of treatment (termed image-guided radiation therapy, IGRT). Both 3D planning and IGRT reduced the high-dose bath to normal tissues by decreasing the size of uncertainty margins used around tumors.

Table

Table

Second, we have been able to decrease the high-dose bath by improving the degree to which the high-dose bath can be “warped” to better align with the shape of the tumor target. Intensity-modulated radiation therapy (IMRT) accomplishes this by targeting a lesion from many directions. The intersection of all these beams can create convex and concave high-dose bath volumes that exquisitely conform to a tumor. The consequence of using many beams is to redistribute normal tissue dose from areas of high dose to low or no dose. In effect, IMRT trades increased volume treated to a lower dose for decreased high-dose bath immediately adjacent to a tumor. This trade-off has allowed radiation oncologists to safely intensify treatment in many disease sites, such as prostate cancer, head and neck cancers, and spine metastases.

Despite our progress, many treatment protocols still require significant uncertainty margins. For example, margins of 3-7 mm are commonly associated with head and neck, breast, and intact prostate cancers, and margins of 12-20 mm are applied for cancers in close proximity to the diaphragm (e.g., liver and lung cancers) or bladder (e.g., cervical and postoperative prostate cancers) (Fig. 2). Because of the strong association of toxicity and the high-dose bath, these large margin expansions into normal tissues limit our ability to intensify treatment. For example, consider postoperative radiation for prostate cancer, which is performed tens of thousands of times across the U.S. each year. There is significant movement and deformation of the prostatectomy bed (the target) between daily doses due to changes in bladder and rectal filling. Thus, when we target a prostatectomy bed, we actually target a volume expanded by a radial uncertainty margin of up to 1 cm. Because volume increases to the third power with linear increases in radius, even small uncertainties can result in large volume changes. In a typical post-prostatectomy case, one might wish to target 250 cc of tissue at risk for harboring microscopic tumor cells. After uncertainty margins are applied, however, the actual volume of normal tissue treated with a high-dose bath is more than 600cc. With more precise targeting, we could spare almost half of the tissue volume in these common cases.

Even in cases where the initial uncertainty margin is small, there may be a much larger effective uncertainty expansion. In oropharynx cancer, for instance, an initial uncertainty expansion of 3-5 mm is commonly used. However, a large percentage of oropharynx cancers will shrink during an 8-week course of radiation, so by the end of a course, a much larger volume of normal tissue will be irradiated than initially prescribed. Whether it be a result of day-to-day variations or changes over weeks, most modern treatment protocols continue to expose unnecessarily large volumes of normal tissue to the high-dose bath.

Back to Top | Article Outline

Adaptive Therapy Minimizes Uncertainty

Exposure of normal tissue to unnecessary high-dose radiation is in large part due to the fact that there is often an extended delay between the start of radiation planning and the start of treatment. In the current radiation treatment planning workflow, a patient receives a diagnostic quality CT scan in the treatment position, a static radiation plan is generated from the scan (a process that may take up to 2 weeks), and the patient returns for that plan to be delivered over the course of therapy which can last up to 8 weeks. Over this extended period of time, the tumor may grow during the planning period or shrink during treatment. Additionally, the position of the tumor can shift in response to bladder filling or breathing.

Historically, there has been no practical way to adapt the treatment plan day-to-day to reflect these changes. However, the integration of high-quality imaging and automated planning software with the linear accelerator (linac) enables updating of a patient's original treatment plan with the most recent data available in only a few minutes and while the patient is on the treatment table. This online adaptive radiation therapy (ART) approach can significantly diminish margins by accounting for anatomical and potentially functional changes that occur from one fraction to the next and over the course of treatment.

Figure

Figure

Back to Top | Article Outline

Potential to Transform Radiation Oncology

Although both CT and MRI-based imaging can potentially be used for online ART, MR-guidance has two important advantages: it can be used to visualize any tumor type and may enable biologic response adaptation to shrink the high-dose bath to the subpopulation that needs it.

CT is inherently limited in its ability to discriminate between tumors and normal soft tissues if these tissues have similar electron densities. Even in sites where density differences are adequate for diagnostic CT, CT-guidance may not be possible given the technical challenges of integrating a diagnostic quality CT scanner into the linac system without interfering with the accuracy of radiation delivery. Diagnostic quality CT scans can only be captured if the scanner is physically separated from the linac, limiting the ability to obtain images when the beam is on.

Figure

Figure

In contrast, recent advances have overcome significant technical barriers, enabling combined MRI and linear accelerator platforms. Such systems are now in development. Because the MR unit on these devices is being developed to produce images of diagnostic quality, a wide range of MR sequences can be used to image tumors throughout the body. Because the MR system is integrated with the linac, it can holistically image both tumors and normal tissues while the beam is on. Thus, MR-guided online adaptive therapy uses daily MR imaging to adapt treatment plans immediately prior to initiation of radiation delivery, which reflects changes in tumor location and shape. The location and shape of the tumor is subsequently monitored by real-time MR images during delivery of that treatment plan. With these technological advances, true anatomic adaptive therapy is becoming a reality.

The second potential advantage of MR-guidance, particularly units with a high-field MRI (1.5 Tesla and up) is that it may enable biologic response-based adaptation. Functional MR sequences have the potential to identify which patients are responding to radiation therapy and which have resistant disease early in the treatment period, even before anatomic changes may be detectable. Periodic functional assessments could potentially be used to identify patients who do not require escalated doses to be cured, sparing them the side effects associated with standard radiation therapy protocols. Conversely, functional assessments may also be able to identify small regions of tumor tissue that may require higher or longer dosing than originally planned to optimize chances of cure. Functional assessments might also be used to monitor normal tissue and identify patients who may have a genetic sensitivity to radiation and would be candidates for alternative therapies. Although functional MR imaging has been available for years, it is only now feasible to integrate the technology into linac systems, enabling cost-effective assessment of tissue fractions throughout treatment. Biologic adaptation of therapy is poised to transform radiation therapy, and MR-linac has the potential to pave the way for other exciting advances in radiomics and radiation biology.

MR-guided ART, while in its early stages, has the potential to be a disruptive technology in cancer care. By combining two of the most powerful tools in the radiation therapy armamentarium—a diagnostic quality MRI and a state-of-the-art linear accelerator—the field is effectively positioned to once again meaningfully shrink the high-dose bath and make its next great advance.

Back to Top | Article Outline

Have Stories Delivered Directly to You

Figure

Figure

Sign up for our free e-newsletter and receive email notifications when each issue becomes available and when other new content is posted online. New to Oncology Times? Visit the publication's website at www.oncology-times.com where you can get free access to archived issues, subscribe to the print magazine, and create your own personal online article collections.

Wolters Kluwer Health, Inc. All rights reserved.
Home  Clinical Resource Center
Current Issue       Search OT
Archives Get OT Enews
Blogs Email us!