10 â€“ Three-Dimensional Conformal Radiotherapy and Intensity-Modulated Radiotherapy

Ping Xia, PhD,
Lei Xing, PhD,
Howard I. Amols, PhD,
C. Clifton Ling, PhD

The success of radiotherapy depends on the radiosensitivity of the particular tumor being treated relative to that of the surrounding normal tissues. The goal in radiotherapy, therefore, is to sufficiently separate the dose-response curves of local tumor control and normal tissue complications. During the past decade, advances in radiologic imaging and computer technology have significantly enhanced our ability to achieve this goal through the development of three-dimensional (3-D) image-based conformal radiotherapy (CRT) and intensity-modulated radiotherapy (IMRT). The implementation of these technologies permits better shaping of the high-dose volume of the radiation treatment so as to better conform to the tumor volume while minimizing the radiation dose delivered to surrounding normal tissue. Success of 3D-CRT and IMRT critically relies on the accurate delineation of the tumor volume, augmented with multiple imaging modalities such as the magnetic resonance imaging (MRI), positron emission computed tomography (PET/CT), and four-dimensional computed tomography (4D-CT). Organ motion, change of the patient's anatomy, and the tumor response to radiotherapy call for the transition from 3D-CRT to 4D-CRT.

IMRT is becoming a mature technology, and is widely applied to many cancer sites. Many treatment-planning comparison studies have demonstrated the clear dosimetric advantages of IMRT.^[1-6] Clinical results from the past decade have shown the improvement of local tumor control and reduction of treatment toxicities for prostate cancer, head and neck cancer, and other types of cancer.^[7-14]

In this chapter, we describe the rationale and processes for 3D-CRT and IMRT, including patient imaging and simulation, treatment planning, dose calculation algorithms, plan evaluation, treatment verification, treatment delivery, and quality assurance (QA) issues. In this discussion, we identify the similarities and differences between 3D-CRT and IMRT, and emphasize those features and benefits unique to IMRT.

Three-Dimensional Conformal Radiotherapy and Intensity-Modulated Radiotherapy Treatment Planning Process

Conventional radiotherapy entails irradiation of the patient from a few beam directions (as the accelerator rotates about the patient), using beam configurations such as that shown in Fig. 10-1, which depicts a five-field isocentric treatment of the prostate. Usually all beams are aimed at a single point denoted as the isocenter that geometrically represents the intersection of the axis of rotation of the linear accelerator gantry head, the collimator assembly, and the treatment couch. Although most often the geometric center of the tumor is purposely positioned at the isocenter, 3D-CRT and IMRT planning do not stringently require that the isocenter be located at the geometric center of the tumor. In many special clinical scenarios (such as in treatment plans for head and neck and breast cancer), the isocenter may be deliberately placed on a specific location, off the geometric center of the tumor. Even though the intervening superficial tissues receive higher radiation doses than does the tumor for each individual beam, the summation of all beams results in a higher dose to the tumor. With conventional radiotherapy, the radiation shape (or aperture) from each beam is manually drawn on the projected two-dimensional (2-D) images, taken from an x-ray machine (referred to as a simulator), which simulates the geometry of a treatment machine (a linear accelerator). With easy access to computed tomography (CT) in modern radiotherapy departments, planning CTs are acquired for most patients undergoing radiotherapy. Sometimes the CT unit, equipped with a flat couch top and alignment laser system, is referred to as a CT simulator. After acquiring a planning CT, the treatment tumor volumes are directly delineated on the CT images, and the radiation portals (or apertures) are designed to conform to the tumor volumes. The difference between conventional radiotherapy and 3D-CRT lies in whether the planning CT is used to define tumor volume and to design treatment portals accordingly. Therefore, 3D-CRT entails more sophisticated shaping of the dose distribution than does conventional radiotherapy because the collimation design (or shaping of fields) and the selection of beam directions are based on 3-D CT images of the patient. These images, projected in a so-called beam's eye view (BEV) format (described more fully later), permit us to select beam directions with short pathways to the tumor and better avoidance of normal tissues. IMRT goes one step beyond 3D-CRT by enabling variations of the radiation intensity within each beam. This intensity modulation can be achieved via several different approaches, including fabrication of complex physical compensators (often fabricated with computer-controlled milling machines) to be placed in the radiation beam between the radiation source and the patient, but more commonly via the use of a multileaf collimator (MLC) capable of dynamic beam delivery or use of multiple static beams sequentially altered in shape by the MLC. The details of these delivery methods are discussed in the third section of this chapter, “Delivery of Intensity-Modulated Treatment.”

First, let us briefly describe the process of 3D-CRT and IMRT, after which we present details of each step in the process. The treatment planning steps for IMRT and 3D-CRT are similar during the initial and final steps, but diverge in the middle. In particular, patient imaging and simulation are identical for both processes. IMRT differs from 3D-CRT in some of the key steps of plan optimization, which we will discuss along the way. The treatment planning process begins with “treatment simulation,” which entails setting up the patient on the CT unit (or a CT simulator) in the treatment position. More frequently, patient CT imaging and simulation is being augmented with MRI, PET, and other functional imaging studies to better define the tumor volume and critical sensitive structures.

To precisely mimic the patient's treatment position, simulation first determines a proper posture for the treatment position, including the use of a proper type of immobilization device to facilitate precise reproduction of patient position during simulation, other image acquisition, and multifraction treatment delivery. Often, during the simulation, a patient-specific body mold is fabricated to fix the treatment posture. Fig. 10-2 shows a patient fixed in a head, neck, and shoulder mask on the treatment table. Subsequently, 3-D images are acquired from which the radiation oncologists can delineate target and nearby sensitive structures. Sometimes medical physicists or dosimetrists assist radiation oncologists to contour common sensitive structures, such as the spinal cord, lungs, and liver. The delineation of anatomic volumes is usually done directly on a computer display of transverse CT images using standard computer graphics options such as a mouse, track ball, light pen, and so on. More often, MRI, PET, or PET/CT images are registered with the planning CT to provide better soft tissue contrast (such as in MRI images) and better physiologic information (such as in PET images).

Once all relevant tissues have been delineated, the radiation oncologist specifies the desired doses (or the prescription doses) to tumor volumes and the limiting doses to normal tissues. From these specifications the medical physicists or dosimetrists then select and adjust beam directions and design shapes and beam intensities so as to best meet these dose criteria. For 3D-CRT plans, the aperture shape and intensity of each beam is manually designed and iteratively adjusted based on the planner's experience and intuition. Because the process is performed via “manual” iteration, the quality of the 3D-CRT plan can be constrained by time and by the fact that x-ray beam intensities within each individual beam are uniform (or, if wedges are added, monotonically variable in one dimension only). Once beam directions, shapes, and intensities are specified, the computer calculates the resulting dose distribution, which one compares with the radiation oncologist's specifications. If there are discrepancies, beam intensities, shapes, and directions are iteratively adjusted (based primarily on the planner's experience and intuition) and dose distributions recalculated. In practice, only a few parameters can really be adjusted, and each iteration is extremely time consuming. Hence the ability to truly optimize a dose distribution with 3D-CRT techniques is often quite limited. This type of iteration is also referred to as forward planning (FP).

IMRT usually incorporates computerized iteration of radiation beams as opposed to the manual optimization just described. In this process, the computer performs a large number of iterations in less time than a human could perform a few iterations. Further, the computer can modulate the dose intensity within each radiation portal, which is typically divided into multiple pencil beams, each on the order of several square millimeters. The combination of sub-beams of different intensities plus a large number of iterations often enables significantly improved dose distributions with IMRT as opposed to 3D-CRT.^[15]

Once a treatment plan or dose distribution has been calculated, the radiation oncologist and planner must evaluate the plan to determine how well it meets the original criteria, or dose constraints. This is typically done via analysis of dose distributions and dose-volume histograms (DVHs). When the treatment plan has been accepted for treatment, all of the planning data on beam configurations and radiation intensities are transferred to the linear accelerator (now often done via computer networking systems and “record-and-verify” [R/V] systems) and patient treatment proceeds using the beams designed during treatment planning. For IMRT, the data transferred includes information on dynamic MLC (DMLC) motion files or on static MLC (SMLC) segmental files required to deliver the desired x-ray intensity profiles. Finally, physicists must perform dosimetric and other QA tasks to verify that all equipment is functioning properly, and that the specifics of the dose prescription and treatment plan are accurately delivered to the patient on a daily basis. In the following sections we discuss each of these treatment planning steps in greater detail.

Patient Setup, Immobilization, and Image Acquisition

The minimization of patient set-up uncertainty is more important in 3D-CRT than in conventional radiotherapy because of the improved conformality of the dose distribution (i.e., quick dose gradient between the boundary of tumor volume and normal tissue). This becomes even more critical in IMRT because IMRT plans often produce even sharper dose gradients at the boundary between the tumor volume and normal tissue.^[16] Thus, immobilization devices and precise patient positioning procedures must be applied throughout the process of image acquisition, simulation, and treatment.