Intensity Modulated Radiation Therapy (IMRT): A Patient-Centered Guide - Jason Lee, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: August 21, 2002
Abstract
Radiation therapy is an important treatment modality in the curative and palliative management of cancer patients. Advances in radiation therapy are occurring at both the biologic and physical level. One significant technological advance is intensity modulated radiation therapy, or IMRT. This two-part article reviews the basic concepts of IMRT, steps involved in treatment planning, methods of IMRT delivery, and some clinical uses of IMRT.
Learning Objectives
After reading this article, patients and students should be able to:
Understand the treatment planning process for IMRT;
Understand the methods of delivering IMRT; and
Understand the clinical uses of IMRT.
IMRT Introduction: Part I
Concept of IMRT
Treatment planning in radiation oncology has undergone major evolution since the first therapeutic use of x-rays in the early 1900's. With a better understanding of cancer biology and normal tissue reactions as well as improved diagnostic imaging tools, radiation oncologists are better able to define the targets for treatment and deliver focused beams of radiation to those targets.
One recent advance began with the introduction of CT scanners, as these images could be transferred to target planning computers where tumor volumes and normal tissues were defined for development of radiation beam arrangements. The process, known as "three-dimensional conformal radiation therapy" (3D-CRT), has proliferated as a method of giving higher tumor dose and minimizing effects on normal tissue.
The development of a treatment plan using 3D-CRT has been performed traditionally with a "forward planning" process, where beam arrangements were tested more or less by trial-and-error, until a satisfactory dose distribution was produced. For complex cases, this process can be very time-consuming because of the number of beam parameters that can be modified.
In contrast, IMRT relies on "inverse treatment planning" and non-uniform radiation exposures to optimize the dose distribution to the target. In inverse treatment planning, the radiation oncologist enters clinical parameters, such as desired dose to the target volume and dose-limits to normal organs, into the targeting computer, which then "back-calculates" from the desired dose-distribution and develops an optimal treatment plan to conform to those parameters (Figure 1).
An IMRT targeting computer also adjusts the intensity of the radiation beam across the field, depending on whether the tumor or sensitive normal tissues lie in the beam path. The availability of inexpensive and powerful computing equipment has made the sophisticated optimization process practical and relatively automated.
In addition, the advent of multileaf collimators (MLC) on modern linear accelerators has permitted the delivery of multiple and complex portal geometries necessary for IMRT. Multileaf collimators can move in and out of the beam portal under precise computer guidance while the radiation is on, thereby generating the desired nonuniform fluence (intensity) patterns that produce a uniform dose to the target.
Figure 1: Inverse Treatment Planning Model
Planning of IMRT
The planning process for IMRT involves several steps:
Position and immobilization,
Patient data acquisition,
Target and normal tissue definition,
Dose prescription and dose limits,
Beam optimization, and
Treatment plan evaluation.
Because IMRT delivers radiation in a precise fashion, patient positioning and immobilization are crucial elements in the planning process. Certain tumor sites, such as head and neck cancer, brain cancer, and prostate cancer, are well-suited for IMRT given their relatively stable position in relation to bony anatomy and close proximity to critical normal structures.
New devices have been implemented for 3D-CRT and IMRT. For head and neck and brain cancers, special reinforced thermoplastic masks hold the patient within a few millimeters. For prostate cancer, some form of body immobilization either using thermoplastics or vacuum cast is implemented. In some centers, the prostate gland can be better localized using either a rectal balloon, transabdominal ultrasound, or intraprostatic radio-opaque markers.
Treatment planning for IMRT relies on CT images of the patient in treatment position. Fixed marks are placed on the patient or the immobilization device and lined up with lasers such that every point within the patient can be localized in three-dimensional space. In the next step, target volumes and normal tissue are outlined on the CT images by the radiation oncologist and dosimetrist.
In IMRT, multiple targets can be specified, each with an unique dose prescription if necessary. One of the greatest strengths of IMRT is its ability to limit dose to normal tissues at risk for radiation damage. These organs commonly include the optic nerve and chiasm, brainstem, parotid glands, spinal cord, kidneys, rectum, bladder, femoral necks, and others depending on the location of the tumor. Figure 2 shows a CT slice of an intra-abdominal tumor and contours of target and normal tissues.
Figure 2: Outlining volumes of interest
Once these volumes are defined, doses are prescribed to targets and dose-limits are assigned to normal tissues. There are a number of methods to specify dose parameters. One method is to assign a dose goal or dose limit to a structure, and a tissue weighting, which reflects the relative importance of constraining to the specified dose.
Typically, the tumor and critical organs receive the highest weighting. In other treatment planning systems, the dose limits incorporate a volume constraint; i.e., the volume of the rectal wall receiving more than 70 Gy should be 40% or less. Following the specification of the dose criteria, the treatment planning computer performs an iterative search to develop an optimized beam configuration and intensity pattern. The dose criteria must be realistic, or the planning computer may fail to produce an acceptable plan.
A number of optimization algorithms have been developed which work to minimize the deviation of the dose distribution from a proposed treatment plan from the desired dose distribution. Once the planning computer produces an optimized plan, a dose distribution is calculated and evaluated by the radiation oncologist. The dose distribution can be assessed subjectively by evaluating an isodose plot, or objectively by reviewing dose volume histograms. If the dose distribution could be improved, the radiation oncologist can modify the dose specifications to "tweak" the plan to his or her liking.
Delivery of IMRT
The delivery of IMRT has been facilitated by the introduction of MLCs. Instead of rectangular edges in the beam aperture of earlier linear accelerators, an MLC consists of narrow leaves which are under computer control and can form custom-shaped portals. During IMRT, the leaves may also slide during radiation exposure, thereby adjusting the intensity of one portion of the beam. The implementation of IMRT may be performed using two distinct methods: multiple fixed gantry positions or a rotation gantry.
With a fixed gantry technique, multiple beam angle and table configurations are chosen which should optimize radiation delivery. An example would be for prostate cancer, where beams commonly enter from anterior, LAO, LPO, RAO, and RPO directions. At each position, radiation delivery occurs through the appropriate portal shape and in an optimal fluence pattern.
The fluence pattern can be adjusted by dynamic (moving) MLC, also known as a "sliding window" technique, or by delivering multiple "segments" of radiation at each gantry position. With the latter technique, the radiation intensity for each segment is constant, while the confluence of the multiple segments produce the modulated fluence profile.
With a rotation gantry technique, the gantry (linear accelerator treatment head) swings around the patient in an arc configuration while the radiation is on, and the field shape and intensity are continually modified. The first available IMRT system divided the target volume into thin slices, and treated each slice sequentially in one arc.
After each arc, the patient is moved horizontally, and the next slice is treated; hence, the term "sequential arc" IMRT. The MLC in this system is a special hardware device that treats a relatively thin slit rather than a full field. Since linear accelerators are now equipped with full field MLC, this adaptation allows full field IMRT with a rotating gantry.
Because the intensity pattern is often complex and may not be feasible with a single arc, the treatment can be divided into multiple arcs, with each arc treating a particular "segment" of the beam profile. This method is known as "intensity modulated arc therapy."
With any implementation of IMRT, new quality assurance measures are necessary to ensure that the intended dose distribution is, in fact, being delivered. Because of the complexity of the treatment, there is no practical method to verify the dose distributions with the patient in place.
This generally requires a "trial run" in some sort of phantom, or container with radiation measuring devices. By exposing the phantom to a patient's set of beam, the actual measured radiation dose can be compared to the calculated dose for the phantom. In addition, the special hardware must undergo quality assurance; i.e. movement of MLCs and movement of the treatment couch.
IMRT Introduction: Part II
Clinical Applications of IMRT
The potential clinical applications of IMRT are broad and continue to expand. Many of the earlier studies have assessed whether IMRT treatment planning is feasible and compared the IMRT plans with conventional dose distributions. Clinical studies of IMRT on actual patients are now beginning to emerge since its first use in 1994.
1. Brain Tumors: IMRT has been used for intracranial tumors (both benign and malignant) and head and neck cancers. Brain lesions may be large, irregular, and solitary, or smaller and multiple; IMRT can address both situations. The dosimetry produced by sequential arc IMRT rivals that of stereotactic radiosurgery. A potential advantage to IMRT is its ability to limit dose to surrounding normal tissues, such as the optic nerve, chiasm, lens, and brainstem, thereby possibly minimizing radiation morbidity.
Figure 3 demonstrated an isodose distribution for a suprasellar lesion, and shows the inward deviation of isodose curves near critical structures. 3D-CRT and IMRT are able to conform to tumor volumes, and has permitted delivery of higher radiation doses to tumor. Trials are underway which are investigating escalating doses to high-grade gliomas and increasing dose-intensity with accelerated fractionation.
Figure 3: Isodose plan for a suprasellar mass
Figure 4: Isodose plan for nasopharynx cancer
2. Head and Neck Cancer: Many of the technique issues for brain tumors also apply to head and neck cancer. There is also great interest in limiting dose to the parotid gland and thus preventing xerostomia, or permanent dry mouth, that occurs with typical head and neck radiotherapy. For example, in Figure 4, there is a cancer of the nasopharynx which has extended into the right oropharynx and upper neck nodes (highlighted in red).
The concave shape tumor around the spinal cord is a typical challenge for radiation oncologists, but an ideal application for IMRT. The resultant IMRT plan is able to direct dose away from the spinal cord and contralateral parotid gland, while maintaining dose uniformity across the gross tumor and subclinical regions at risk (highlighted in purple).
IMRT can potentially simplify treatment planning as there is no need for matching electron fields or multiple conedown simulations. In early clinical trials, patients have reported less dry mouth and shown greater salivary flow with parotid sparing techniques compared with conventional radiotherapy. As the results from clinical trials emerge, it will be important to ensure that the reduction in high-dose volume does not compromise tumor control.
3. Prostate Cancer: Radiation therapy has been a mainstay of localized prostate cancer therapy for several decades. Technological advances in treatment planning have permitted higher doses to the prostate and better shielding of the rectum and bladder. IMRT is particularly suited for dose escalation in prostate cancer, as this technique provides avoidance of rectum, bladder, and femoral necks to minimize potential morbidity.
There is good rationale for dose escalation in prostate cancer, based on evidence from randomized trials from M.D. Anderson Cancer Center and a phase I trial from Memorial Sloan Kettering Cancer Center. By increasing dose from 70 to 78 Gy, intermediate-risk prostate cancers were more likely to be controlled. In addition, doses of 81 Gy resulted in a 7% positive biopsy rate compared with 45% or more with lower doses. Many centers with both 3D CRT and IMRT capabilities have adopted IMRT as the preferred treatment planning method for prostate cancer.
Limitations of IMRT
The implementation of IMRT can be a tedious task, even for radiation oncology facilities equipped with 3D-CRT. IMRT requires special hardware and physics expertise as well as some amount of re-training for radiation therapists. Some of the potential advantages and disadvantages of IMRT are shown in the table below. With greater experience, it is possible for IMRT to become a very efficient treatment technique.
Because of its recent introduction to clinical radiation oncology, many more years of study will be necessary to determine the true therapeutic impact of IMRT on tumor control, toxicity, and patient survival.
Advantages
Conformal dose distribution around tumor
Avoidance of critical structures and less local toxicity
Computer-generated optimization
Disadvantages
Equipment costs higher
Special immobilization required
Treatment time often longer
Additional quality assurance necessary
Learning curve can be steep
