Last Modified: August 21, 2002
Authors: Jason Lee MD,1 Indra J. Das PhD,1 Shiao Y. Woo MD,2
Walter Grant PhD,2 Bin S. Teh MD,2 J. Kam Chiu MD,2
E. Brian Butler MD2
Affiliations: Abramson Cancer Center of the University of Pennsylvania1 Department of Radiation Oncology, The Methodist Hospital, Baylor College of Medicine2
Implementation of IMRT
Treatment systems for IMRT have become available with the introduction of multileaf collimators (MLC). Instead of rectangular collimator jaws found in early radiation treatment units, an MLC consists of multiple narrow leaves which are under computer control and allow custom-shaped beam apertures without fabricated blocks. The width of the leaves on a typical MLC projects between 5 mm to 12.5 mm at isocenter, depending on the manufacturer. The characteristics of MLC depends on the manufacturers' design, and has been extensively reported. (15-17) During IMRT delivery, the leaves of the MLC are adjusted while the beam is on to modify the fluence of radiation across the portal. The treatment portal is broken into smaller deliverable portals, called beamlets, which can be delivery by the MLC. The physical equivalent of MLC-based IMRT would involve the fabrication of a custom-milled lead alloy block of varying thickness to match the fluence profile necessary for each portal -- a very cumbersome task.
The implementation of IMRT may be performed using two distinct methods: multiple fixed gantry positions or a rotating gantry. With fixed gantry therapy, multiple beam angle and table configurations are chosen which should optimize radiation delivery. For example, for prostate cancer, the beam may enter from anterior, LAO, LPO, RAO, and RPO directions. A greater number of gantry positions increases the flexibility of dose modulation and improves dose distributions up to a certain point. With a rotation gantry technique, treatment is delivered from multiple arc rotations.
1. Fixed Gantry, "Sliding Window:" At each gantry position, opposing leaves of the MLC are operated in dynamic mode (ie, in motion while the radiation is on) at a certain speed and distance to produce a portion of the beam aperture. A number of algorithms have been developed to convert the fluence profile into an MLC motion sequence. A hypothetical example of beam intensity profiles for an LAO portal used in prostate cancer is shown in Figure 4-1. The use of this technique for prostate cancer has been popularized at Memorial Sloan Kettering Cancer Center (New York, NY). (18)
Figure 4-1: Schematic of beam intensity profiles of an LAO field for prostate cancer
2. Fixed Gantry, "Multisegment:" For each gantry position, the portal is divided into multiple "segments" with individual monitor unit settings. While the radiation intensity for each segment is constant, the confluence of the segments produce the modulated fluence profile. Verification of the dose for each stationary segment is possible. At the University of Michigan Medical Center (Ann Arbor, MI), treatment for head and neck cancer consists of five gantry angles and one to four segments at each gantry position. (19)
3. Rotating Gantry, "Sequential Arc:" With the sequential arc technique, the treatment is delivered in multiple axial slices. Each slice of the target region is treated with a thin-slit MLC during gantry rotation. After each rotation, the patient is moved in the horizontal direction so the next slice can be treated. The Peacock system is a form of this technique and has been successfully used at various institutions including the Methodist Hospital (Houston, TX) for brain, head and neck, and prostate cancer patients. (20) This process can be time-consuming since only one slice is treated at a time.
4. Rotating Gantry, "Intensity-Modulated Arc Therapy:" A stationary patient may be treated by a full-field MLC portal with a series of superimposed intensity-modulated arcs. Each arc treats a certain "segment" of the beam profile, and cumulative dose deposit fluence from all the arcs form the desired dose-distribution. (21) This is sometimes known as "double-dynamic" MLC IMRT, since both the gantry and MLC leaves are in motion.
5. Rotating Gantry, "Tomotherapy:" A generalization of a sequential arc technique is known as tomotherapy, where both the gantry and patient are in motion, much like a helical CT scanner. (22) This technique is under development at the University of Wisconsin (Madison, WI). A CT scanner on the same gantry would allow real-time dose verification. This application is a fusion of two technologies- CT and linear accelerators.
Commercially Available Treatment Systems
Since the introduction of the first IMRT system in 1994, a number of commercial systems have become available. This report discusses a few of these products below, but by no means is this a comprehensive review.
1. Peacock (NOMOS Corporation, Sewickley, PA)
The first FDA-approved IMRT system known as the Peacock was introduced in 1994 and is in wide clinical use today. Over 1,000 patients have been treated with the Peacock system at The Methodist Hospital (Houston, TX). The Peacock takes advantage of the sequential arc technique of IMRT. (23) The Peacock delivers radiation through a special collimator device called the Multileaf Intensity Modulating Collimator (MIMiC) which attaches to the gantry head of a standard linear accelerator (Figure 4-7). The MIMiC has opposing rows of 20 tungsten leaves, each of which projects to a 1 ? 1 cm2 field at isocenter in the open position (Figure 4-8). The MIMiC is also known as a binary system, which allows only a 1 ? 1 cm2 pixel by each leaf. A single arc with the MIMiC treats a 2 cm length; longer fields are treated with sequential arcs with precision linear indexing of the couch. The leaves of the MIMiC are pneumatically-driven by computer control while the gantry is rotating, thereby creating the desired intensity-modulated pattern. For each individual patient, the information for gantry and leaf position is stored on a 3.5 inch floppy disk and inserted into the on-board computer of the MIMiC for treatment. During gantry motion, the pneumatic positioning of the MIMiC leaves produces loud "clack-clack-clack" sounds [Audio], and patients are offered ear plugs to reduce this noise.
Figure 4-7: Peacock attachment on linear accelerator
Figure 4-8: MIMiC collimator unit accelerator
The information for gantry and leaf position is obtained from the inverse treatment planning computer, after the dose-distribution has been approved by the radiation oncologist. The Peacock delivers radiation through sequential 270o arcs via a portal approximately 20 ? 2 cm or 20 ? 4 cm. For every 5o segment (total 54 segments) of the arc, the radiation intensity can be varied through each of the 40 leafs in 10% increments from completely closed (0%) to completely open (100%). The leaf intensity pattern is generated by the inverse planning computer, which uses an optimization algorithm known as "simulated annealing" to conform to the dose-constraints set by the treating physician. The result of this spatially nonuniform radiation exposure is a highly conformal and relatively uniform dose administration to the target.
2. Varian Medical Systems, Inc. (Palo Alto, CA)
The linear accelerators produced by Varian and other manufacturers use full-field MLC to deliver IMRT, rather than sequential arcs as in the Peacock system. The premium MLC marketed by Varian contains 60 pairs of leaves, forming a full 40 ? 40 cm field. The central leaves measure 5 mm at isocenter and can be moved with a precision of 0.2 mm, thereby creating the high resolution beam apertures necessary for precision IMRT. With Varian's inverse treatment planning software, this equipment is suitable for "step and shoot" or "sliding window" IMRT. Varian equipment and software have undergone extensive development, testing, and clinical use at Stanford University, Memorial Sloan Kettering Medical Center, the Medical College of Virginia, and other academic centers.
The inverse treatment system (HeliosTM) uses a sophistated optimization algorithm that provides plans for varying beam energies and coplanar or noncoplanar configurations. The optimization process considers beamlets measuring 2.5 ? 5-10 mm, depending on leaf width, and typically 30 voxels per cm3 for target and normal tissue regions. The dose calculation algorithm factors in head scatter and leaf transmission when generating the leaf motion sequences. The result is highly accurate radiation delivery compared with the calculated dose distributions.
3. Other Inverse Planning Systems
The MLC and IMRT planning system are separate entities and can be obtained separately. Most MLC's constructed today are designed to take advantage of IMRT. A number of inverse planning systems are available from companies with existing 3D-CRT planning software. NOMOS, makers of the Peacock, has developed the CORVUS system which uses dose-volume constraint specifications and can drive both the MIMiC and full-field MLC's. PLATO Complete (Nucletron Corporation, Columbia, MD) and Helax-TMS 5.0 (MDS Nordion, Ontario, Canada) are other inverse planning/IMRT systems which have gained 510(k) clearance by the FDA for marketing in the U.S. Newer software incorporates sophisticated models which can calculate effects of head scatter, thus producing more accurate dose distributions, particularly in low-dose off-axis regions. The treatment plans generated by these packages can be delivered by MLC-equipped linear accelerators.
A concern with computer-controlled treatment is the validation of the radiation delivery especially when using beamlets; in other words, are the tumor and surrounding tissues receiving the planned dose? Verification of dosimetry is basically a comparison of measured versus calculated dose for a particular plan. With conventional radiation therapy, a well-aligned portal film and a dosimeter measurement are generally sufficient reassurance. In IMRT, there is no practical method to verify spatial dose-intensity gradients with the patient in place. A number of potential obstacles are present in IMRT which may result in a deviation from the planned dose: 1) patient setup and organ motion; 2) couch movement in sequential arc therapy; 3) accuracy of leaf movement on dynamic MLC; and 4) radiation leakage through MLC leaves and beam characteristics from longer beam-on times. Aslo for segmented treatment, the radiation is delivered in beamlets for a very short time. This raises serious issues regarding dose linearity as described by Das et al. (24 ) Several techniques have been devised which can validate that the intended dose-distributions are, in fact, being delivered.
1. "Square Box" Phantom: Assessing the changes in optical density of radiographic film when exposed to ionizing radiation is a familiar method of measuring dose. A square box phantom holds multiple sheets of film at regular spacing, and is then irradiated with the IMRT plan. The film densities are measured with an optical scanner and compared with predicted planar dose distributions for the phantom. Alternatively, the phantom can be fitted with a matrix of individual dosimeters (TLD or diodes) to obtain direct dose measurements. (25) For prostate cancer treatment verification, the phantom can be covered with a tissue-equivalent shell to more closely approximate the contour of a human pelvis (Figure 4-9).
Figure 4-9: Pelvic film phantom
2. Anthropomorphic Phantom: Unlike square phantoms, a humanoid phantom has the advantage of performing quality assurance in targets shaped and similar to treated patients. An anthropomorphic phantom known as RANDO® (The Phantom Laboratory, Salem, NY) has been created which corresponds to the average male and female body shapes and densities. RANDO® can accomodate film and individual dosimeters, though there is less freedom with placement and film must be shaped to fit. One study comparing the patient-optimized plans calculated for the humanoid phantom and the measured dose distributions in the phantom revealed that dose discrepancies do occur, but most are less than 3%. (26)
3. Electronic Portal Dose Imaging (EPID): Many newer linear accelerators can be equipped with an on-line electronic portal imaging device. By operating this device in a manner which integrates the captured radiation over the entire beam area, the resultant portal image could be compared with a computer-constructed portal dose image. This system would have to account for the degradation in image quality with longer distances from the radiation source and the effect of scattered dose. Additional problems include the inhomogeneity present in the patient's body, whereas calculations are performed as a homogeneous phantom.
4. BANGTM (MGS Research Inc., Guilford, CT) Gel Imaging: This is a newer dosimetry system where a special gel compound known as BANGTM is exposed to IMRT or stereotactic treatment. The ionizing radiation produces a polymerization reaction in the gel which can be imaged by MRI. A full 3D dose distribution can be obtained, with a resolution as small as a few cubic millimeters. (27) The need to obtain an MRI scan for the gel limits the practicality of this method; however, an optical measuring system is under development.
During the early commissioning of an IMRT system, quality assurance procedures for each new patient are advisable to verify proper dose delivery. Additionally, the precision of linear couch movement in sequential arc IMRT, as in Peacock treatment, requires inspection as table indexing errors as little as 1 mm can lead to dose inhomogeneity of 10%. (11) A facility may elect to perform quality assurance measures for individual patient plans on a less frequent basis once additional experience and a higher comfort level is gained with IMRT.
The potential clinical applications of IMRT are broad and continue to expand. While IMRT is feasible for any disease site treated with conventional radiation therapy (with the exception of total body irradiation), IMRT has been implemented where benefits are most likely to occur. IMRT is particularly well-suited for curative treatment of intracranial tumors, head and neck tumors, and prostate tumors, where target location is fairly stable, normal tissue toxicity is of major concern, and a dose-response has been shown for local control.
New radiation therapy technology can be evaluated by various "levels" of evidence, similar to the safety, efficacy, and effectiveness data required for pharmaceuticals. The first level is whether IMRT can produce the desired dose distributions, and the evidence from the discussion above would support this. The second level is whether IMRT is safe for patients, particularly in the setting of dose escalation. The third level is whether IMRT can result in cure or local control of tumors, and lastly, whether IMRT is superior in terms of cure rates and toxicity compared with conventional technology. Studies on IMRT for various clinical sites are reviewed briefly below.
Because of the stability of the cranium and the sensitivity of numerous intracranial structures, much of the early IMRT activity focused on brain and head and neck tumors. IMRT has been used for malignant and benign tumors, and in adult and pediatric populations. Table 5-1 lists the major avoidance structures such as the orbits, optic nerve and chiasm, brainstem, and temporal lobes, along with suggested dose limits. Brain lesions may be large, irregular, and solitary, or smaller and multiple. IMRT can address both situations. The dosimetry using Peacock-based IMRT and stereotactic radiosurgical techniques for various clinical scenarios is often comparable. Peacock and LINAC-based stereotactic plans were compared for simulated intracranial targets using RTOG criteria of homogeneity (how uniform the dose is over the target) index and conformality (how well dose is limited to the target) index. (28) For regular-shaped lesions, they were similar, but for large irregular lesions >4 cm, the Peacock plan was actually more conformal and homogeneous.
Figure 5-1: Peacock isodose curves for a suprasellar mass
For lesions that have been traditionally treated with conventional radiation therapy, IMRT may result in reduced morbidity by virtue of its ability to avoid surrounding normal tissues. Examples include craniopharyngiomas and pituitary adenomas which are adjacent to temporal lobes, optic nerve, and optic chiasm. An axial isodose distribution using Peacock for a suprasellar lesion (red) is shown in Figure 5-1, and demonstrates the inward deviation of the isodose curves near these critical structures (optic chiasm and brainstem in green).
For high-grade gliomas, little progress has been made in improving cure rates for the past several years. Relatively few studies have reported the experience using 3D-CRT or IMRT for these brain tumors. Some centers are investigating dose-escalation beyond the current standard dose of 60 Gy. At the University of Michigan, doses of 90 Gy have been delivered to glioblastoma multiforme patients and escalation continues. (5) At the Methodist Hospital (Houston, TX), a protocol is investigating the safety and efficacy of accelerated IMRT given as daily fractions of 5 Gy to GTV and 2 Gy to CTV over 10 fractions.
|Normal Structure||Dose Threshold (cGy)||Normal Structure||Dose Threshold (cGy)|
|Brainstem||5,000||Ipsilateral Parotid Gland||3,500|
|Spinal Cord||4,000||Contralateral Parotid Gland||2,500|
Table 5-1: Radiation Thresholds for Head and Neck and Intracranial Tumor Treatment
Head and Neck Tumors
Many of the immobilization and avoidance issues for brain tumors also apply to head and neck cancers. In addition, there is great interest in sparing parotid gland salivary function and preventing xerostomia (dry mouth), a common complication from conventional irradiation of head and neck cancers. The geometry of head and neck cancers is often complex, and efforts are also directed towards reducing dose to adjacent structures such as the mandible and spinal cord. Typical dose-limits to critical structures are shown in Table 5-1. As the results from clinical studies emerge, it will be important to ensure that the reduction in high-dose volumes does not compromise locoregional control.
Figure 5-2 shows an isodose curve of a nasopharynx cancer that has extended to the right oropharynx and upper neck nodes. The gross visible tumor (highlighted in red) on CT scan has a concave shape surrounding the spinal cord, a common situation in advanced head and neck cancer and typical challenge for radiation oncologists. The Peacock plan is able to direct high-dose regions away from the spinal cord, mandible, and opposite parotid gland, while maintaining dose-uniformity across the gross tumor and subclinical regions at risk (highlighted in purple). Furthermore, this treatment plan simplifies matters as it obviates the need for matching electron boosts once spinal tolerance has been exceeded.
Figure 5-2: Peacock plan for nasopharynx cancer
Parotid-sparing radiation therapy for head and neck cancer has been investigated by the University of Michigan using 3D-CRT and more recently, IMRT. (29) IMRT is implemented using multisegmented static fields, with the contralateral parotid gland as an avoidance structure. A comparison of treatment plans versus reconstructed conventional three-field plans in 15 patients showed better target dose conformity and a dose-reduction in contralateral parotid of about two-thirds. Salivary flow was measured in 88 patients before and after conformal parotid-sparing radiation therapy. By limiting the mean dose received by the parotid gland to less than 26 Gy, salivary flow was preserved in the majority of cases, whereas high doses resulted in little recovery of salivary production. Compared with standard irradiation, patients reported less dry mouth with conformal parotid-sparing techniques. (19) In a recent study of patterns of failure following conformal irradiation, the 2-year local control was 79%, with none recurring near the contralateral parotid gland. However, 4 of 58 patients experienced relapses in the superior neck and retropharyngeal nodes near the ipsilateral parotid, stimulating the investigators to increase doses to these regions. (30)
IMRT has the capability of incorporating boost treatment simultaneously within larger field irradiation. (31,32) This technique of head and neck radiation may provide a higher biologically equivalent dose while reducing treatment time and maintaining acute toxicity at acceptable levels. These techniques are further discussed below.
Locoregional recurrence in head and neck cancer is not an uncommon situation. Management options for these patients tend to be limited and designed with palliative intent. Because of normal tissue tolerance, conventional re-irradiation is often not possible, and studies have reported on stereotactic techniques, hyperfractionated radiation, and brachytherapy. At the Methodist Hospital, ten patients with recurrent disease (previous doses 50-67 Gy) underwent Peacock-based IMRT for an additional 16-66 Gy. All patients achieved palliation of local symptoms, and half of the patients have had no progression of local disease. (20)
Radiation therapy has a long history in the treatment of prostate cancer. A series of technological advances in the past 20 years have improved the targeting of the prostate and shielding of rectum and bladder. This has lead radiation oncologists to embark on dose-escalation trials for prostate cancer in attempt to improve tumor control. (33) However, higher doses can induce more rectal and bladder toxicity. IMRT has been used for prostate cancer over the past several years to allow dose-escalation without equivalent increase in toxicity. Because of the narrow prescribing margins (often <10 mm) between target and organs at risk (rectum and bladder), novel immobilization and location methods are necessary. One approach is the insertion of a rectal balloon filled with air which presses the prostate against the pubis, and stabilizes the gland during therapy. Other methods include transabdominal ultrasound location of the prostate gland, or placement of metal seeds within the gland to serve as internal fiducial markers.
Memorial Sloan Kettering Cancer Center has been conducting a dose-escalation study for prostate cancer since the late 1980's. Their initial treatment plan consisted of six static 3D-planned coplanar fields, and a boost above 72 Gy was performed with 6-8 lateral and posterior-lateral fields. An IMRT-based plan using the "sliding window" method was developed for the boost, and quickly became the technique for the entire course of therapy. (18,34) The dose-distributions produced by IMRT were even superior to those produced by a 6-field 3D-CRT technique. The preliminary results of this trial using conventional 3D-CRT are encouraging, and support further dose-escalation.35 In 743 patients, the probability of a PSA nadir to 1.0 ng/ml or less and the probability of PSA control were higher in the higher dose arms. Among the selected patients who underwent post-treatment biopsies, fewer patients in the highest dose arm (81 Gy) were positive for tumor cells. In addition, grade 3 rectal and urinary toxicity was <1% for all patients, a risk similar to or less than that associated with conventional doses.
|Dose Level||Patients Treated||Patients Biopsied||Negative for Tumor Cells||Tumor Cells with Radiation Effect||Positive for Tumor Cells|
Table 5-2: Biopsy Results in Prostate Cancer Dose-Escalation Trial
An alternative to dose-escalation is accelerated treatment with IMRT. Investigators at the Cleveland Clinic have tested the safety and efficacy of hypofractionated IMRT to 70 Gy in 28 fractions. (36) This has the advantage of shorter treatment times (5.5 weeks for 70 Gy vs. 9 weeks for 81 Gy), with similar biologic effect. Fifty-one patients were treated with 5 static intensity-modulated fields generated by CORVUS inverse planning. Approximately 70% of patients experienced grade 1 acute rectal and bladder toxicity, and up to 20% experienced grade 2 toxicity, which is more frequent than that generally seen with conventional fractionation. If this proves to be an effective and tolerable regimen, the shortened duration of treatment may enhance patient convenience.
Prostate cancer patients are treated with the Peacock system at the Methodist Hospital (Houston, TX). A typical isodose plan for a prostate patient is shown in Figure 5-3. Patients are prescribed 70 Gy in 35 fractions to the periphery of the prostate gland and seminal vesicles. This corresponds to a mean dose to the prostate and seminal vesicles of approximately 76 Gy and 74 Gy respectively, and less than 35 Gy to the bladder and rectum. A toxicity analysis was performed on 50 patients and compared with conventionally-treated patients (Table 5-3). Although this is a retrospective comparison, there is a suggestion that patients treated with IMRT may have less acute toxicity. (37)
Figure 5-3: Peacock plan for prostate cancer
|GU Toxicity by RTOG Grade||GI Toxicity by RTOG Grade|
Table 5-3: Acute Radiation Toxicity in Prostate Cancer Treatment
There has been recent interest in using IMRT for tangential treatment of breast cancer patients. The goals of breast IMRT are twofold: 1) reduce dose to the ipsilateral lung, contralateral breast, and if left-sided, heart and coronary vessels; and 2) for institutions who treat the internal mammary nodal (IMN) chain, incorporate IMNs while maintaining dose-limits on normal structures. A dosimetric study was performed at Memorial Sloan Kettering Cancer Center, using IMRT implemented with the sliding window technique. (38) Doses to critical normal structures, including the contralateral breast and coronary arteries represented by the left anterior cardiac wall, were significantly reduced.
A second study compared conventional tangential irradiation with static field IMRT in 20 breast patients. (39) The method of intensity modulation involved an initial field prescribed to approximately 90% of the presscribed dose, and a second segment of radiation which supplemented underdosed regions of the breast. Dose homogeneity across the breast could be improved from the "hot spots" of 22% seen with conventional therapy. More importantly, the 90% of the IMN chain received 4500 cGy, whereas 50% of the IMN chain received 3500 cGy with conventional planning, without increased dose to the underlying normal organs.
Simultaneous Boost Therapy
Standard radiation fractionation schemes deliver higher doses to areas of greatest disease burden; i.e., gross tumor receives 60-70 Gy and subclinical regions receive 45-50 Gy. This has conventionally been performed with successive phases of radiation, the first delivering lower doses to larger volumes and subsequent phases directed at smaller target volumes. In head and neck cancer, this scheme is known as "shrinking field" technique, and may involve several conedowns. Tumor control in head and neck cancer is also related to length of treatment time, with protracted courses resulting in lower control rates. (40) Accelerated fractionation with twice-a-day radiation has been proposed to counteract repopulation of tumor clonogens, and shown to be superior compared with standard fractionation using a concomitant boost technique in a randomized trial. (41)
IMRT has the ability to incorporate a concomitant boost volume within larger volume treatment. This technique is known as a simultaneous integrated boost (SIB) or simultaneous modulated accelerated radiation therapy (SMART). (31) For a typical head and neck case, the gross tumor at the primary and nodal sites may be designated target 1, and prescribed 220-250 cGy per day. The surrounding subclinical regions adjacent to the primary tumor and regional nodes may be designated target 2, and prescribed 180-200 cGy per day. Radiobiologically, the resultant dose-fractionation is likely to be similar to a typical dose of 7000 cGy in 7 weeks. A comparison of the standard fractionation, accelerated fractionation, and the SMART boost is shown in Table 5-4.
|Schedule||Dose-Fractionation||Total Dose (cGy)||Overall Duration|
|Standard fractionation||200 cGy x 35 fractions QD||7,000||7 weeks|
|Hyperfractionated Radiotherapy||120 cGy x fractions BID||7,440||6 weeks + 1 day|
MGH Split Course
200 cGy x 35 fractions QD 7/wk
|Concomitant Boost (MDACC)||180 cGy x 30 fractions QAM
150 cGy x 12 fractions QPM
|SMART Boost||240 cGy x 25 fractions (GTV)
200 cGy x 25 fractions (CTV)
|CAIR = continuous accelerated radiation therapy; MGH = Massachusetts General Hospital; CHART = continuous hyperfractionated accelerated radiation therapy; CHARTWEL = CHART weekend-less; MDACC = M.D. Anderson Cancer Center; SMART = simultaneous modulated accelerated radiation therapy|
Table 5-4: Accelerated Fractionation Schemas
The potential advantages of an integrated boost are that a single plan can be used for the entire treatment, and fractionation to gross disease is accelerated. In addition, the dose distribution is more conformal, since the boost fields do not "violate" previously treated regions that were intended to receive subclinical doses. The downside is that this treatment can be more toxic, as the daily dose to mucosal regions is higher.
The experience with the SMART boost in the first 20 Peacock-treated head and neck patients was recently reported. (31) Gross tumor and subclinical disease are commonly prescribed 240 cGy and 200 cGy per day, respectively, or 6000 cGy over 5 weeks to gross disease. The mean doses to target regions are higher- 6440 and 5440 cGy. Grade 3 mucosal toxicity was observe in 80% of patients, resulting in the need for IV fluids or tube feeding in many patients. Late toxicity has not been observed, as the mean doses to critical structures have been kept well below tolerance. Local disease has been controlled in 90% of patients, although two patients developed lung metastases.
The implementation of IMRT can be a complex and tedious task, even for radiation oncology facilities with 3D-CRT already in use. Dynamic MLC must be installed either on existing linear accelerators if feasible or as a package with a new accelerator purchase. If a CT-simulator is not present, a mechanism for obtaining patient CT imaging data must be available. Additionally, MRI images are helpful and image fusion software should be available. Planning of IMRT and control of dynamic MLC requires new computer software and a thorough understanding of the planning and treatment process by physics staff. Quality assurance measures must be adopted to ensure IMRT dose delivery is accurate. At Memorial Sloan-Kettering Cancer Center, a full 14-step procedure had been in place once an optimized IMRT plan had been developed.18 Based on prior experience gained over the past several years, a facility should be able to commission an IMRT system in a 3-4 month period. (42)
The planning of IMRT requires a paradigm shift for dosimetrists and physicians. For facilities without existing 3D-CRT planning, the task of image segmentation, or tumor and normal structure contouring, will need to be incorporated into the planning process. While some software may perform some of these tasks automatically or nearly so, the time requirement for contouring alone may take between 30 minutes or more than a hour, depending on the site of disease, organs at risk, and number of CT slices. Instead of manually testing beam arrangements and weightings, physicians will need to formulate tumor dose goals and normal structure dose-volume constraints, and allow computers to achieve the solution. For some common tumor sites such as the prostate gland, relatively standardized IMRT solutions are in place, thus streamlining the dosimetry operation.
For radiation therapists, additional training will be required regarding precision patient setup, new immobilization devices, and use of new hardware and software. While computer-controlled MLC has obviated the need for fabrication of blocks and for the therapist to enter the treatment room between fields to change blocks, treatment times may not be shorter and are, for some tumor sites, often longer. At the Methodist Hospital (Houston, TX), the total treatment time of a typical prostate cancer patient with three table indices on a Peacock system is 15 to 20 minutes. (43) Even with sliding window technique and IMAT, the treatment time is between 20 and 40 minutes. For a head and neck patient with 5 to 7 table indices and sometimes an additional conventional field for low neck irradiation, the treatment time can be substantially longer. However, initial and boost fields can be combined into a single plan, and supplemental electron fields are often unnecessary. With more automated equipment, higher dose rate accelerators, and greater experience, patient throughput is likely to improve.
The costs of any new radiation treatment relate to the fixed costs of medical equipment, the resources required to provide the treatment, and the patient-related costs of travel and self-care during a course of therapy. In addition, downstream costs related to the potential morbidity and treatment for cancer recurrences may require consideration if these differ among treatment options. (44) Depending on the facility, acquisition of a 3D planning system and an MLC device may represent a substantial investment. These costs differ depending on the manufacturer and existing equipment setup, and this discussion is beyond the scope of this review. Patient-related indirect costs are similar for prostate cancer treatment, unless dose-escalation is pursued and additional treatment time is required. For head and neck cancer, the incorporation of a SMART boost saves patients 2 weeks of treatment time, without the need for twice-a-day visits.
In one economic study, the Medicare allowable charges for conventional treatment and IMRT were compared for prostate cancer and head and neck cancer. (43) As shown in Table 6-1, the charges for IMRT for prostate cancer were about 33% higher than conventional therapy. For head and neck cancer, the charges with a SMART boost were 20% lower than conventional therapy, and 33% less than twice-a-day treatment. The charge formulas will fluctuate across various radiation oncology facilities, but this elementary analysis demonstrates IMRT need not be drastically more expensive than conventional treatment.
|Radiation Schedule||Professional Charges||Technical Charges||Total Medicare Allowable Charges|
|Head and Neck Cancer
Concomitant Boost Radiation
|IMRT = intensity-modulated radiation therapy; SMART = simultaneous modulated accelerated radiation therapy|
Table 6-1: Medicare Allowable Charges for Various Radiation Schedules43
Limitations of IMRT
3D-CRT and IMRT are certainly major advances in the physical delivery of radiation therapy to a target. This technology is relatively new and still being perfected. IMRT is an inefficient method of delivering radiation, since a large number of radiation monitor units are wasted in the process of achieving the desired beam fluence. In the Peacock system, a single arc can deliver 240 cGy to a target at a dose rate of 300 cGy/min. Higher fraction sizes would require repetitive arcs at the same table index, or a higher output accelerator. Efficiency of dose delivery at the present time is only 5 to 10%, whereas over 90% of output is wasted.
As a result of the increased monitor units required for IMRT, it is possible that radiation scatter and whole-body doses will increase. The dose at 10 cm from the target was measured at 2.5% of the prescribed dose. (45) In order to produce the conformal dose distributions while sparing critical surrounding organs, IMRT spreads out dose to noncritical normal tissue. The volume of normal tissue receiving low levels of radiation dose can be significantly higher than with conventional radiation therapy. The biologic effects of widespread scattered and low dose radiation are not well understood and require more study. At the same time, dose inhomogeneity within the target volume may expand as a consequence of normal tissue avoidance, and this effect on tumor control should be evaluated carefully. (46)
There are a number of issues related to human and tumor biology which IMRT is unable to overcome. IMRT is poorly suited for tumors in moving organs, such as the lung or adjacent sites whose positions vary with respiration. A number of techniques are being explored to compensate for respiratory motion, but these are far from routine use. Even the small movement that may occur in the prostate as a result of bladder or rectal content can result in significant deviations from desired dose deposition given the ever slimmer margins prescribed. Inhomogeneity corrections also play an important role in inverse planning and delivery. Currrently, all plans are done using a homogeneous phantom. An additional problem is related to defining the target, as the borders of many cancers are difficult to distinguish from normal tissue, even with CT images or MRI co-registration. (47) No imaging study can detect microscopic infiltration of cancer cells, and close attention must be paid to whether marginal or out-of-field recurrences may result from higher doses to tighter margins. Lastly, the assumption that higher doses of radiation beyond conventional doses of 60 to 70 Gy leads to better tumor control has been confirmed for prostate cancer, (48) but for few other cancers. Doses as high as 80 or 90 Gy have been unable to control gliomas, (5) so more radiation and elaborate dose-distributions will clearly be insufficient for certain cancers.
Will more patients be cured with IMRT? Higher tumor doses may be feasible while limiting normal tissue toxicity which, in theory, may improve local control of disease. Ultimately, the improved local control and reduction in radiation morbidity should result in better patient outcomes. Unfortunately there has been a relative paucity of rigorous scientific data to support the clinical benefits of IMRT; evaluation of the safety and efficacy of IMRT has been largely based on short-term comparison with historical data from conventionally treated patients. However, the early experience has been very encouraging, and many academic centers continue to push this technology to even further limits.
IMRT is a promising treatment technique for a variety of clinical situations. The implementation of IMRT relies on concepts of inverse treatment planning and computer-controlled modulation of radiation intensity. Issues regarding commissioning and quality assurance of IMRT systems have been addressed. Dosimetric studies have demonstrated improvements in conformality to target regions, and the crucial ability to avoid organs at risk for radiation injury. Clinical studies have demonstrated a reduction in acute and long-term morbidity, particularly with rectal and bladder toxicity from prostate irradiation and xerostomia. 3D-CRT and IMRT may allow safe dose-escalation with potential improvement in local control. Early results in tumor control with IMRT are encouraging; however, more long-term data are needed to confirm these benefits over conventional radiation therapy.
AcknowledgmentsThe authors thank the radiation therapists, dosimetrists, and staff at the Baylor College of Medicine Methodist Hospital for their kind assistance with the preparation of this review.
Yu CX. Intensity-modulated arc therapy with dynamic multileaf collimation: an alternative to tomotherapy. Phys Med Biol 1995;40:1435-1449.
Mackie TR, Balog J, Ruchala K, et al. Tomotherapy. Sem Radiat Oncol 1999;9(1):108-117.
Carol MP. PeacockÔ : a system for planning and rotational delivery of intensity-modulated field. Int J Imaging Systems Tech 1995;6:56-61.
Das IJ, Kase JR, Tello VM. Dosimetric accuracy at low monitor unit settings. Br J Radiol 1991;64:808-811.
Low DA, Gerber RL, Mutic S, Purdy JA. Phantoms for IMRT dose distribution measurement and treatment verification. Int J Radiat Oncol Biol Phys 1998;40(5):1231-5.
Tsai JS, Wazer DE, Ling MN, et al. Dosimetric verification of the dynamic intensity-modulated radiation therapy of 92 patients. Int J Radiat Oncol Biol Phys 1998;40(5):1213-30.
Low DA, Dempsey JF, Venkatesan R, et al. Evaluation of polymer gels and MRI as a 3-D dosimeter for intensity-modulated radiation therapy. Med Phys 1999;26(8):1542-51.
Woo SY, Grant WH 3rd, Bellezza D, et al. A comparison of intensity modulated conformal therapy with a conventional external beam stereotactic radiosurgery system for the treatment of single and multiple intracranial lesions. Int J Radiat Oncol Biol Phys 1996;35(3):593-7.
Eisbruch A, Dawson LA, Kim HM, et al. Conformal and intensity modulated irradiation of head and neck cancer: the potential for improved target irradiation, salivary gland function, and quality of life. Acta Otorhinolaryngol Belg 1999;53(3):271-5.
Dawson LA, Anzai Y, Marsh L, et al. Patterns of local-regional recurrence following parotid-sparing conformal and segmental intensity-modulated radiotherapy for head and neck cancer. Int J Radiat Oncol Biol Phys 2000;46(5):1117-26.
Butler EB, Teh BS, Grant WH III, et al. SMART (simultaneous modulated accelerated radiation therapy) boost: a new accelerated fractionation schedule for the treatment of head and neck cancer with intensity modulated radiotherapy. Int J Radiat Oncol Biol Phys 1999;45(1):21-32.
Wu Q, Manning M, Schmidt-Ullrich R, Mohan R. The potential for sparing of parotids and escalation of biologically effective bose with intensity-modulated radiation treatments of head and neck cancers: a treatment design study. Int J Radiat Oncol Biol Phys 2000;46(1):195.
Zelefsky MJ, Leibel SA, Kutcher GJ, Fuks Z. Three-dimensional conformal radiotherapy and dose escalation: where do we stand? Semin Radiat Oncol 1998;8(2):107-14.
Ling CC, Burman C, Chui CS, et al. Conformal radiation treatment of prostate cancer using inversely-planned intensity-modulated photon beams produced with dynamic multileaf collimation. Int J Radiat Oncol Biol Phys 1996 Jul 1;35(4):721-30.
Zelefsky MJ, Leibel SA, Gaudin PB, et al. Dose escalation with three-dimensional conformal radiation therapy affects the outcome in prostate cancer. Int J Radiat Oncol Biol 1998;41(3):491-500.
Mohan DS, Kupelian PA, Willoughby TR. Short-course intensity-modulated radiotherapy for localized prostate cancer with daily transabdominal ultrasound localization of the prostate gland. Int J Radiat Oncol Biol Phys 2000;46(3):575-80.
Teh BS, Uhl BM, Augspurger ME, et al. Intensity modulated radiotherapy (IMRT) for localized prostate cancer: preliminary results of acute toxicity compared to conventional and six field approach (abstr). Int J Radiat Oncol Biol Phys 1998;42(Suppl. 1):219.
Hong L, Hunt M, Chui C, et al. Intensity-modulated tangential breast irradiation of the intact breast. Int J Radiat Oncol Biol Phys 1999;44(5):1155-64.
Lo Y-C, Yasuda G, Fitzgerald TJ, Urie MM. Intensity modulation for breast treatment using static multi-leaf collimators. Int J Radiat Oncol Biol Phys 2000;46(1):187-194.
Robertson C, Roberston A, Hendry JH, et al. Similar decreases in local tumor control are calculated for treatment protraction and for interruptions in the radiotherapy of carcinoma of the larynx in four centers. Int J Radiat Oncol Biol Phys 1998;40:319-329.
Fu KK, Pajak TJ, Spencer SA, et al. A Radiation Therapy Oncology Group (RTOG) phase III randomized study to compare hyperfractionation and two variants of accelerated fractionation to standard fractionation radiotherapy for head and neck squamous cell carcinomas: preliminary results of RTOG 9003 [Abstract]. Int J Radiat Oncol Biol Phys 1999;45(3 Suppl.):145.
Grant WH III. Commissioning and quality assurance of an IMRT system. In: Sternick ES (ed): The Theory and Practice of Intensity Modulated Radiation Therapy. Madison, WI: Advanced Medical Publishing, 1997, pp 121-126.
Grant WH III, Woo SY. Clinical and financial issues for intensity-modulated radiation therapy delivery. Sem Radiat Oncol 1999;9(1):99-1107.
Perez CA, Michalski J, Ballard S, et al. Cost benefit of emerging technology in localized carcinoma of the prostate. Int J Radiat Oncol Biol Phys 1997;39(4):875-883.
Mutic S, Low DA. Whole-body dose from tomotherapy delivery. Int J Radiat Oncol Biol Phys 1998;42:229-232.
Goitein M, Niemierko A. Intensity modulated therapy and inhomogeneous dose to the tumor: a note of caution. Int J Radiat Oncol Biol Phys 1996;36(2):519-522.
Glatstein E. Dr. Strange (high) tech or how I learned to stop worrying and love my MLC/3D treatment planning, stereotactic linac. Int J Radiat Oncol Biol Phys 1999;45(5):1097-101.
Pollack A, Zagars GK, Smith LG, et al. Preliminary results of a randomized dose-escalation study comparing 70 Gy to 78 Gy for the treatment of prostate cancer [Abstract]. Int J Radiat Oncol Biol Phys 1999;45(3 Suppl.):146.