CHALLENGES IN THE CLINICAL APPLICATION OF ADVANCED TECHNOLOGIES TO REDUCE RADIATION-ASSOCIATED NORMAL TISSUE INJURY

Transcription

1 doi: /j.ijrobp Int. J. Radiation Oncology Biol. Phys., Vol. 69, No. 1, pp. 4 12, 2007 Copyright Ó 2007 Elsevier Inc. Printed in the USA. All rights reserved /07/$ see front matter CRITICAL REVIEW CHALLENGES IN THE CLINICAL APPLICATION OF ADVANCED TECHNOLOGIES TO REDUCE RADIATION-ASSOCIATED NORMAL TISSUE INJURY LAWRENCE B. MARKS, M.D.,* AND JINLI MA, M.D.* y *Department of Radiation Oncology, Duke University Medical Center, Durham, NC; and y Department of Radiation Oncology, Fudan University Cancer Hospital, Shanghai, China INTRODUCTION There is rapidly growing interest in applying advanced treatment planning/delivery technologies to improve the therapeutic ratio of external beam radiation therapy (RT). The evolution from two-dimensional (2D) to three-dimensional (3D) conformal planning, and now to intensity-modulated radiation therapy (IMRT) and image guidance (i.e., imageguided radiation therapy [IGRT]), have altered our approach to minimizing normal tissue injury. A broad review of these techniques is presented, with a primary focus on the potential pitfalls of applying these new approaches to minimize normal tissue toxicities. It is critical that physicians understand the power and limitations of these technologies so that they can be optimally exploited to reduce the risks to normal tissues and improve the therapeutic ratio of radiation therapy. BACKGROUND 2D VS. 3D In the pre-3d era, physicians prescribed RT without full understanding of the 3D doses and volumes. Clinically safe techniques were defined based largely on doses, fractionation, and field sizes/orientations. Experienced physicians relied on their judgment, experience, and qualitative assessment of proposed treatment fields, often without objective quantitative metrics. Three-dimensional planning tools provide images depicting the beam path through the patient (1, 2). This allows beam orientations/shapes to be more optimally selected to improve the therapeutic ratio (i.e., spare normal tissue or increasing target coverage) (1). Alternative treatment beams, and groups of beams, can now be quantitatively compared based on 3D dose and volume data. Three-dimensional planning is a major advance in radiation oncology because it provides fundamental anatomic and dosimetric information heretofore not available. For any particular case, clinicians often disagree on what should be considered the target volume and what dose to deliver. However, all physicians agree that all of the targeted tissue should be irradiated. Three-dimensional planning facilitates this. The 3D dose distribution is challenging to comprehend. This has led to the development of representative surrogates of the 3D dose distribution that are easier to understand. The commonly used dose volume histogram (DVH) takes the 3D dose volume data, discards all spatial information, and represents it as a 2D graph (3). Because such DVHs can also be difficult to interpret, it is common to extract representative figures of merit (e.g., mean dose, or the percent of an organ receiving in excess of a given dose) from the dose distribution (4). The potential shortcomings of DVHs and figures of merit are discussed later. Normal tissue dose volume limits in the 3D era One of the great goals of 3D planning was that it would increase our knowledge of the dose volume dependence of normal tissue injury (5). In large measure this has become true. Table 1 summarizes several sites in which 3D dose volume data have been successfully related to normal tissue outcomes. Figure 1 illustrates some of the data for several specific organs. Such quantitative information is now widely used during the treatment planning process. It is often difficult to determine which dosimetric parameter best predicts for organ dysfunction because there is often a strong correlation between the different parameters. In the lung, for example, the correlation coefficient between the competing dosimetric parameters (e.g., mean lung dose, Reprint requests to: Lawrence B. Marks, M.D., Department of Radiation Oncology, Duke University Medical Center, Box 3085 DUMC, Durham, NC Tel: (919) ; Fax: (919) ; lawrence.marks@duke.edu Dr. Marks has received honoraria from Varian Medical Systems, as well as grant support from the NIH, the Lance Armstrong Foundation, and the Department of Defense. 4 Acknowledgment Thanks to Jane Hoppenworth and Donna Wimberley for their assistance in manuscript preparation. Received Feb 13, 2007, and in revised form May 7, Accepted for publication May 7, 2007.

2 Reducing normal tissue injury d L. B. MARKS AND J. MA 5 Table 1. Representative organs where DVH-based metrics have been successfully related to the risk/severity of normal tissue injury Organ DVH-based metrics* End point Representative references Lung Mean dose, V5, V13, V20, V25, V30, NTCP Radiation pneumonitis (6 16) Parotid Mean dose, V15, V30, V45 Salivary flow (17 19) Rectum V40, V45, V50, V60, V65, V70, Dmax >75 Gy Proctitis/bleeding (20 24) Liver NTCP, mean dose, V30 Liver disease (RILD) (25 29) Heart Mean dose, V45, V50, NTCP Heart disease (RIHD) (30 33) Brain V12, Dmean20 Edema/breakdown of the blood-brain barrier, necrosis (34, 35) Abbreviations: DVH, dose volume histogram; RILD and RIHD = radiation-induced liver and heart disease; Dmean20 = the mean dose (in Gray) in a specified volume of 20 cm 3 that was given the highest dose; NTCP = normal tissue complication probability computed using a model based on the entire three-dimensional dose distribution. * Vx is the percent of the organ receiving $X Gy. For several organs, different dose thresholds (e.g., V20, V30) have been suggested to correlate with injury, as shown. percentage of lung volume receiving $20 Gy) is >0.9 (8, 9, 13, 36). This high correlation is likely due to the use of similar RT techniques within the population of patients studied. When the RT techniques become more diverse (e.g., with IMRT), the correlation between these parameters may become less strong. Thus it is not clear if parameters that were predictive in the 3D era will remain predictive with IMRT. For the parotid, the mean dose and partial volume thresholds are highly correlated with each other (37), and thus mean parotid dose is commonly used (17, 19). 3D vs. IMRT With conventional 3D treatment planning, the entire target is typically encompassed within each beam. Further, each RT beam typically delivers a relatively homogeneous dose throughout the target. Thus significant RT dose is typically received by tissues in the shadow of the target as seen in the beam s-eye view. The selection of the beam orientation is critical with 3D planning (38 40). Wedges can be added to the beam to modify the intensity profile, but are relatively simple and provide only uniform and monotonic modulation of the beam s intensity (e.g., the entire anterior part of a lateral beam is given less intensity than the posterior aspect). The 3D dose distribution resulting from the sum of multiple beams with relatively-uniform intensity is typically a convex polygon (41). With IMRT, each portion of the beam is modulated to provide a unique intensity to each region of the tumor (42 45). Thus the dose to normal tissue within the path of the beam can be adjusted. The purposefully nonuniform intensities from several beam orientations combine to deliver the desired 3D dose (41, 44 48). The IMRT provides the planner/physician flexibility to modify the 3D dose distribution (41, 44). Extremely conformal, and purposefully nonuniform, distributions can be generated. The IMRT appears particularly useful for concave targets. Because the entire target does not need to be irradiated by each beam, the number of feasible beam orientations increases. Further, the selection of beam orientations may be less critical for IMRT than for 3D, because the portion of the IMRT-planned beam that shadows a critical normal tissue can be given a low intensity. Conversely, IMRT may require a larger number of beam orientations than 3D to yield the desired dose distribution (45) (Table 2). Fortunately, advanced RT delivery systems readily facilitate this. The impact of IMRT on normal tissue doses The IMRT affords the planner with increased flexibility to more conformally sculpt the RT dose distribution, particularly for concave targets, and hence reduce the risks to normal tissues. For example, encouraging results have been reported for parotid sparing (17, 18). However, there are several limitations of IMRT that deserve highlighting (45, 49). Incorporation of dose volume constraints from the 3D era into IMRT optimization Current dose volume guidelines were mostly derived in the 3D era, typically from patients irradiated with a modest number of beams, with the normal tissue receiving approximately 2 Gy/fraction. With IMRT, the fraction size delivered to the normal tissue is usually variable, and is typically not 2 Gy per fraction. It is not clear if 3D-era data will be applicable in the IMRT era. This uncertainty is illustrated for the endpoint of pneumonitis. Historically, the whole lung was thought to be able to tolerate approximately Gy using conventional fractionation, without chemotherapy (50, 51). Several studies from the early 3D era successfully related pneumonitis rates to the volume of lung exposed to Gy (i.e., V20 or V30) (6, 8). When using conventional beam orientations (e.g., 40 Gy opposed anteroposterior, plus Gy opposed oblique), the V20 or V30 approximates the volume of lung included in any of the RT beams. The obvious method to reduce the V20 is to treat from additional orientations. More recent studies, some using a larger number of beam orientations, have implicated the volume of lung receiving 5 13 Gy (14, 15). Indeed, it is possible that V20 is predictive within the confines of the conventional beam orientations. The use of a larger number of beam orientations may not reduce the normal tissue risks, but rather just alter the parameters that are predictive for toxicity. The IMRT generally requires more beam orientations, and distributes dose to a larger volume of lung. A recent report from Memorial Sloan-Kettering Cancer Center (MSKCC) notes a higher

3 6 I. J. Radiation Oncology d Biology d Physics Volume 69, Number 1, 2007 Table 2. Comparisons between 3D-CRT and IMRT Number of treatment beam orientations Selection of beam orientation Shape of resultant therapeutic dose volume Uniformity of the dose to the target 3D-CRT Typically 2 5 Critical Typically convex Relatively homogeneous IMRT Typically more Less critical Convex or concave Can be homogeneous, but often is heterogeneous Abbreviations: 3D-CRT = three-dimensional conformal radiation therapy; IMRT = intensity-modulated radiation therapy. Fig. 1. The association between dose volume histogram based figures of merit and rates of normal tissue toxicity for several organs. (a) Radiation pneumonitis vs. mean lung dose (6, 10, 12, 13, 16); (b) hepatic dysfunction vs. equivalent mean liver dose. Adapted with permission from Dawson et al. (26); and (c) relative salivary flow vs. the mean doses to the left and right parotid glands. Adapted with permission from Chao et al. (17). rate of pneumonitis in patients treated with IMRT vs. conformal 3D beams (52). Thus care must be taken when applying 3D predictors in the IMRT era. Conversely, a recent study from the M.D. Anderson Cancer Center (MDACC) suggests lower radiation pneumonitis rates for IMRT vs. 3D (53). Parallel vs. series organs and the neighborhood effect It is often useful to categorize normal tissue architecture as either parallel or series (analogous to electrical circuits), because it assists with the selection/creation of models/methods used to analyze DVHs (54). For series organs, injury to any part of the organ may render it dysfunctional (e.g., a bowel stricture affects food mobility throughout the intestine). Conversely, damage to one region of a parallel organ may not significantly alter global function (e.g., injury to a small region of liver, kidney, or lung is often inconsequential). However, there are data to suggest that the dose delivered to a portion of an organ may have an effect on the RT response to the neighboring portion of that organ, for both parallel and series organs. This is well illustrated by a series of elegant studies by van der Kogel s laboratory (55). When a 4-mm length of spinal cord was irradiated, the dose to cause paralysis in 50% of the animals (ED50) was 87.8 Gy. If the adjacent cord (both superior and inferior to the 4-mm central area) was given a modest dose of 4 Gy, the ED50 for the central irradiated cord dropped to 61.2 Gy. If the cord on only one side (i.e., superior or inferior, but not both) was irradiated, an intermediate value of 68.6 Gy for ED50 was obtained. A dose of 18 Gy delivered to both sides of the 4-mm zone reduced the ED50 to 30.9 Gy. Similar data are reported from MSKCC for rectal injury the risk of injury was dependent on the volume of rectum exposed to approximately 70 Gy and 45 Gy (20, 23). Therefore, increasing the volume of normal tissue exposed to low doses of RT may alter the ability of parts of an organ to tolerate focal areas of high dose. These data challenge the classical segregation of organs into parallel vs. series. Because IMRT tends to increase the volume of normal tissue exposed to a low dose, the widespread use of IMRT may have unexpected sequences on the rate of RT-induced normal tissue injury. Dosimetric The IMRT increases the ability to redistribute the dose within the patient, typically not eliminating incidental normal tissue exposure. For a constant dose delivered to the tumor, the integral dose delivered to the patient is approximately similar for IMRT vs. 3D planning (56 59). The qualitative explanation for this is shown in Fig. 2. For an idealized spherical target within a spherical patient, increasing the number of beams will proportionally increase the volume of normal tissue irradiated, yet similarly proportionally decrease the dose to the irradiated

4 Reducing normal tissue injury d L. B. MARKS AND J. MA 7 Fig. 2. The integral dose to the patient is roughly independent of the beam orientation used, assuming a roughly spherical tumor within a roughly cylindrical patient (a c). The integral dose can be reduced by orienting the treatment beam along the long axis of the target (d), or by treating along orientations where the patient s thickness is lowest (e) (As previously noted by others, including Michael Goiten). Repeating this exercise on a reduced patient leads one to conclude that the integral dose to each patient shell is approximately constant (f). Fig. 3. Schematic illustrating the impact of the dose response curve for regional injury on how the redistribution of dose by intensitymodulated radiation therapy might influence the risk to normal tissues. In the example shown, reducing the lung within the high dose region (with a plateau), in exchange for an increased volume exposed to a low dose, with no change in integral dose, may increase integral injury if there is no threshold for injury (solid line). Conversely, if there is a threshold dose for regional injury (dashed line), then increasing the volume exposed to a low dose might be beneficial. volume. The integral dose can be reduced by selectively using beam orientations that pass through portions of the patient with lesser separation (e.g., the anteroposterior vs. the lateral orientation for tumors in the abdomen or pelvis) or oriented along the long-axis of the target. Nevertheless, because most tumors are relatively spherical in shape, and most patients are relatively cylindrical in shape, the integral dose with different treatment beam orientations should be relatively constant. Indeed, quantitative data based on competing plans generated for patients with cancer of the brain, prostate, and lung, suggest that integral dose is relatively insensitive to the RT technique used (58, 60). Further, if we pretend to remove the outer shell of the patient, and repeat the exercise shown in Fig. 2a e on a reduced patient, one again finds that the integral dose is roughly constant for the different approaches. By repeating this process, one must conclude that the integral dose to each shell within the patient stays roughly constant, regardless of the treatment technique (Fig. 2f). Thus, the average dose gradient within the entire patient is also roughly constant. Thus, the power of IMRT to redistribute dose within the patient appears to be somewhat limited. The RT doses to portions of normal tissue within these shells can be moved, but usually not eliminated (60). As increased knowledge of the dosimetric predictors of normal tissue injury within the context of IMRT become available, IMRT will allow the planner to redistribute dose to minimize the normal tissue risks. Because the dose response curve for regional injury is not fully known for many organs, it is unclear how to best exploit IMRT to reduce normal tissue risks. For example, if there is a plateau in such a dose response curve within the clinical dose range, trading a high dose to a small volume for a larger volume at a lower dose (with a constant integral dose) may increase global injury (Fig. 3). Conversely, if there is a threshold to regional injury, such a trade may reduce the risk of injury. Peer-to-peer quality assurance for 3D vs. IMRT: The application of prior knowledge Both 3D and IMRT start with segmentation of a 3D image set to define targets and normal tissues (Fig. 4). For 3D planning, beams and apertures are defined by the planner. Typically, beam orientations resemble conventional beams (48). The experienced physician can apply their prior knowledge to review the beams and 3D dose distribution. The physician can review the beam and doses, and instinctively know if the treatment is safe (i.e., not too large of a risk for normal

5 8 I. J. Radiation Oncology d Biology d Physics Volume 69, Number 1, 2007 Fig. 4. The work flow for conventional three-dimensional (3D) planning vs. intensity-modulated radiation therapy. Both are based on 3D anatomy. For conventional 3D, the one can apply prior knowledge to assess the acceptability of a proposed treatment. tissues, not too small and thus at risk for a marginal miss). This application of prior knowledge is an independent quality assurance check on 3D-planned beams. This quality assurance step is applied during planning and throughout treatment, as beam s-eye view and portal images are viewed, respectively. These images provide a rapid method to visualize the treated volume. Conversely, IMRT planning is a DVH-driven optimization process. The output from an IMRT optimization is a series of intensity profiles, a 3D dose distribution, and DVHs. This certainly includes a wealth of quantitative data to evaluate the overall treatment plan. However, intensity profiles are not yet intuitive to most users, IMRT dose distributions are often more complex than those derived from conventional 3D planning, and DVHs have limitations (discussed in the following section). The quality assurance provided by prior knowledge may be somewhat reduced with IMRT. For decades, chart rounds has been a group quality assurance meeting where the radiation oncology team (e.g., physicians, dosimetrists, physicists) can rapidly check each-other s work (61). Traditionally, a central component of this process is the review of simulation/beam s-eye view/portal images. This provides a rapid and accurate measure of the treated volume. The prescribed and treated films can be readily compared. The chart rounds process needs to undergo an evolution in the IMRT era. A comprehensive review of the segmented planning images, the 3D dose distribution, and the localization images taken on the machine, is a time-consuming but necessary process for all patients (45, 62, 63). The physician should particularly review the image segmentation as this is the fundamental bias of modern treatment planning; yet, in many ways is the most uncertain and variable. It may be most useful to have chart rounds limited to specific anatomic sites where image segmentation is reviewed in detail. However, this might not be practical for small departments/practices. Quality assurance from the radiation therapists With conventional planning, either 2D or 3D, the light projected onto the patient by the shaped field, as seen by the therapist at the machine, generally provides a good representation of the treated volume. Therapists can provide real-time quality assurance regarding what was being treated. Gross errors in setup may be detected by visual inspection of the treatment light field. For example, if a block is inadvertently omitted from a brain field, and the light field is encroaching on the eye, this is readily apparent to the therapists. Similarly, for patients with grossly evident tumors, the light field typically fully encompasses the involved area. Therapists are trained to look at the light field to assess if the treated area matches the prescription (62). With IMRT, there may be little relationship between the setup marks on the skin and the treated volume. This component of quality assurance has thus been reduced. Not forgetting the simple things With the recent advances, it is tempting to rely heavily on the technology to minimize normal tissue risks. However, relatively simple techniques to improve the therapeutic ratio of RT should not be abandoned. For example, increasing the physical distance between targets and normal tissues is a powerful way to reduce morbidity. Patient positioning remains critically important for our patients (64). For example, lateral decubital positioning and belly boards can be an effective manner to displace small bowel. When such nonstandard positioning is used, care must be taken to assure that treatment can be set up in a reproducible fashion. Placing bubble wrap within areas of skin folds, or positioning patients to reduce such skin folds, can help to reduce acute skin reactions. It is often useful for physicians to observe patients in their immobilization device, in the treatment position. Scars, drain sites, skin folds, and other superficial landmarks are often not readily discernible on treatment planning images. This clinical information is often lost if the physician is not intimately involved in the immobilization/planning process. As we become more focused on performing radiation planning on digital images, it is conceivable that patients can proceed through immobilization, planning, and treatment, without intimate bedside physician involvement. The degree of respiratory excursion of a superficial target (e.g., the breast) can often be readily discerned by watching the patient breathe. Similarly, the degree of intrathoracic motion, and the location of opacified bowel can often be readily discerned via fluoroscopy. These maneuvers are typically easier, faster, and cheaper than more sophisticated imaging techniques. Over-reliance on DVHs Dose volume histograms are a 2D representation of the 3D dose distribution. Their utility is limited by the ability to accurately segment the images. Multiple studies have demonstrated marked inter-physician differences in image segmentation. Further, DVHs discard all spatial information and inherently assume that all regions of an organ are equally important. This is not true for many organs. For example, the nephrons located closer to the renal hilum have slightly more

6 Reducing normal tissue injury d L. B. MARKS AND J. MA 9 urine concentrating ability than do the cortical nephrons (65). Different regions of the femoral head are particularly important for weight bearing (66, 67). The lower lobe of the lung may be more important than the upper lobe in the genesis of radiation pneumonitis (8, 68, 69). In the diseased lung, areas of emphysema have less perfusion, and are thus less functionally important, than other regions. Regional differences in salivary function have been noted within different parts of the parotid gland (70). Similarly, xerostomia may result from dose to both the parotid gland as well as the minor salivary glands within the oral cavity (71). Specific subregions of the pharyngeal musculature may be particularly important for development of dysphagia following treatment of the cancers of the head and neck (72). In routine radiation oncology practice, it is not clear if these spatial variations are meaningful enough to incorporate into the optimization process. The DVHs can be weighted to consider these functional differences, to derive dose function histograms (73). One could then perform IMRT optimization based on these dose function histograms (74 76). The DVHs are only as good as is the anatomic information provided by routine imaging (e.g., computed tomography [CT]). Computed tomography may not accurately reflect the underlying anatomy. The esophagus illustrates this point. On CT, the esophageal circumference is usually seen to vary at different axial levels. This reflects anatomic folds and peristaltic motion that is not well visualized by CT. In reality, the circumference of the esophagus is relatively constant through the chest. In an in-depth analysis of this phenomenon, we demonstrated that DVHs that consider these anatomic realities (and assign the same importance of each axial level to the DVH) are better correlated with esophageal toxicity than are traditional DVHs that do not consider this anatomic reality (77). Alternative approaches to considering the longitudinal and circumferential character of the esophageal dose have been suggested (78). The DVHs of organs at risk are typically considered individually; interactions between organs are usually not considered. There are some data to suggest that such interactions may be important. In a series of elegant experiments in rats, portions of the lung, heart, or both were irradiated with protons. Doses to the heart and lung both appeared to be important determinants of RT-induced shortness of breath, assessed by respiratory rate (79, 80). Several studies in humans note a higher rate of pneumonitis in patients with lower lobe tumors (i.e., those likely with a higher cardiac dose) (8, 68, 69), and hence suggest that there might be an interaction between lung and heart irradiation in the development of lung toxicity (31). Potential problems that may arise from the over reliance on DVHs apply to both IMRT and conventional 3D conformal planning. Becoming too conformal? The unmasking of our anatomic uncertainties The International Commission on Radiation Units and Measurements has defined several volumes to better clarify the uncertainties reflected in the definition of the treated Fig. 5. The relationship of gross tumor volume (GTV), clinical target volume (CTV), internal target volume (ITV), and planning target volume (PTV, see text). Advances in (for example) positron emission tomography (PET) imaging, respiratory/motion-gating/control, and in-room imaging can help to reduce the uncertainties in the definition of the GTV, ITV, and PTV, respectively. An understanding of tumor biology, anatomy, and patterns of spread are needed to define the CTV. volume (81, 82). This is a useful construct and is outlined in Fig. 5. The successive enlargements from GTV (gross tumor volume) to CTV (clinical target volume), from CTV to ITV (internal target volume), and from ITV to PTV (planning target volume) are meant to reflect our uncertainties regarding microscopic extension, internal motion, and setup accuracy, respectively. As shown, newer technologies have increased our ability to define, or to reduce these uncertainties. For example, positron emission tomography imaging is assisting with definition of the GTV (83 85). Internal motion is being better quantified with four-dimensional CT, and the ITV can be reduced with respiratory gating or breath-hold techniques (86). Imaging of the patient on the treatment machine increases our knowledge of setup uncertainties (87, 88). Margins for setup uncertainty can be customized for individual patients (89 91), or markedly reduced if imaging is performed and reviewed before each RT session (89, 92). Alternative localization methods can provide similar improvements. Expansion of the treated volume beyond the GTV reflects these various uncertainties. Historically, physician ignorance regarding the disease extent (a biologic uncertainty) was typically believed to be of smaller magnitude than the other uncertainties. Physicians may have derived a level of reassurance that the tumor was included within the irradiated volume because RT field sizes usually far exceeded the size of the known cancer, due largely to the nonbiologic uncertainties. Whether or not this reassurance was justified is debatable. Advances in (for example) imaging of the patient in the treatment room, and tumor tracking, allow a marked reduction in the ITV and PTV margins. Therefore, the physician is now forced to be more specific with GTV and CTV definition. This requires that the physician accurately interpret the diagnostic imaging, understand the limitations and strengths of

7 10 I. J. Radiation Oncology d Biology d Physics Volume 69, Number 1, 2007 the diagnostic tools, and know the likely patterns of tumor spread. Physician uncertainties regarding these issues remain. Care must be taken not to compromise tumor coverage by making the RT fields too tight. The physician needs to balance the risk of normal tissue injury with tumor control. For example, a study from Israel reported an increased rate of local failure with conformal techniques for patients with orbital tumors, likely reflecting marginal misses (93). There presently are no imaging modalities to accurately define the CTV. The CTV definition is based on physician understanding of the tumor biology, anatomic considerations, and biologic/imaging uncertainties (94, 95). Summary Modern radiation planning and delivery tools provide the physician with increased quantitative 3D dose and volume information, as well as increased flexibility to alter these dose volume values. Physicians must understand the power and limitations of these technologies so that they can be optimally exploited to reduce the risks to normal tissues and improve the therapeutic ratio of radiation therapy. REFERENCES 1. Lichter AS, Ten Haken RK. Three-dimensional treatment planning and conformal radiation dose delivery. In: De Vita V, Hellman S, Rosenberg SA, editors. Important advances in oncology. Philadelphia: JB Lippincott Publishers; p Photon Treatment Planning Collaborative Working Group. Three-dimensional display in planning radiation therapy: A clinical perspective. Int J Radiat Oncol Biol Phys 1991;21: Drzymala RE, Mohan R, Brewster L, et al. Dose-volume histograms. Int J Radiat Oncol Biol Phys 1991;21: Kutcher GJ, Burman C, Brewster L, et al. Histogram reduction method for calculating complication probabilities for threedimensional treatment planning evaluations. Int J Radiat Oncol Biol Phys 1991;21: Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to therapeutic irradiation. Int J Radiat Oncol Biol Phys 1991;21: Hernando ML, Marks LB, Bentel GC, et al. Radiation-induced pulmonary toxicity: A dose-volume histogram analysis in 201 patients with lung cancer. Int J Radiat Oncol Biol Phys 2001; 51: Kwa SL, Lebesque JV, Theuws JC, et al. Radiation pneumonitis as a function of mean lung dose: An analysis of pooled data of 540 patients. Int J Radiat Oncol Biol Phys 1998;42: Graham MV, Purdy JA, Emami B, et al. Clinical dose-volume histogram analysis for pneumonitis after 3D treatment for non-small cell lung cancer (NSCLC). Int J Radiat Oncol Biol Phys 1999;45: Seppenwoolde Y, Lebesque JV. Partial irradiation of the lung. Semin Radiat Oncol 2001;11: Rancati T, Ceresoli GL, Gagliardi G, et al. Factors predicting radiation pneumonitis in lung cancer patients: A retrospective study. Radiother Oncol 2003;67: Rodrigues G, Lock M, D Souza D, et al. Prediction of radiation pneumonitis by dose-volume histogram parameters in lung cancer A systematic review. Radiother Oncol 2004;71: Kim TH, Cho KH, Pyo HR, et al. Dose-volumetric parameters for predicting severe radiation pneumonitis after three-dimensional conformal radiation therapy for lung cancer. Radiology 2005;235: Kong FM, Hayman JA, Griffith KA, et al. Final toxicity results of a radiation-dose escalation study in patients with non-smallcell lung cancer (NSCLC): Predictors for radiation pneumonitis and fibrosis. Int J Radiat Oncol Biol Phys 2006;65: Wang SL, Liao Z, Vaporciyan AA, et al. Investigation of clinical and dosimetric factors associated with postoperative pulmonary complications in esophageal cancer patients treated with concurrent chemoradiotherapy followed by surgery. Int J Radiat Oncol Biol Phys 2006;64: Uno T, Isobe K, Kawakami H, et al. Dose-volume factors predicting radiation pneumonitis in patients receiving salvage radiotherapy for postlobectomy locoregional recurrent nonsmall-cell lung cancer. Int J Clin Oncol 2006;11: Yorke ED, Jackson A, Rosenzweig KE, et al. Dose-volume factors contributing to the incidence of radiation pneumonitis in non-small-cell lung cancer patients treated with three-dimensional conformal radiation therapy. Int J Radiat Oncol Biol Phys 2002;54: Chao KS, Deasy JO, Markman J, et al. A prospective study of salivary function sparing in patients with head-and-neck cancers receiving intensity-modulated or three-dimensional radiation therapy: Initial results. Int J Radiat Oncol Biol Phys 2001;49: Eisbruch A, Ship JA, Dawson LA, et al. Salivary gland sparing and improved target irradiation by conformal and intensity modulated irradiation of head and neck cancer. World J Surg 2003;27: Chao KS. Protection of salivary function by intensity-modulated radiation therapy in patients with head and neck cancer. Semin Radiat Oncol 2002;12: Jackson A. Partial irradiation of the rectum. Semin Radiat Oncol 2001;11: Wachter S, Gerstner N, Goldner G, et al. Rectal sequelae after conformal radiotherapy of prostate cancer: Dose-volume histograms as predictive factors. Radiother Oncol 2001;59: Fiorino C, Sanguineti G, Cozzarini C, et al. Rectal dose-volume constraints in high-dose radiotherapy of localized prostate cancer. Int J Radiat Oncol Biol Phys 2003;57: Jackson A, Skwarchuk MW, Zelefsky MJ, et al. Late rectal bleeding after conformal radiotherapy of prostate cancer. II. Volume effects and dose-volume histograms. Int J Radiat Oncol Biol Phys 2001;49: Boersma LJ, van den Brink M, Bruce AM, et al. Estimation of the incidence of late bladder and rectum complications after high-dose (70 78 GY) conformal radiotherapy for prostate cancer, using dose-volume histograms. Int J Radiat Oncol Biol Phys 1998;41: Lawrence TS, Ten Haken RK, Kessler ML, et al. The use of 3-D dose volume analysis to predict radiation hepatitis. Int J Radiat Oncol Biol Phys 1992;23: Dawson LA, Normolle D, Balter JM, et al. Analysis of radiation-induced liver disease using the Lyman NTCP model. Int J Radiat Oncol Biol Phys 2002;53: Dawson LA, Ten Haken RK. Partial volume tolerance of the liver to radiation. Semin Radiat Oncol 2005;15: Cheng JC, Liu HS, Wu JK, et al. Inclusion of biological factors in parallel-architecture normal-tissue complication probability model for radiation-induced liver disease. Int J Radiat Oncol Biol Phys 2005;62: Liang SX, Zhu XD, Xu ZY, et al. Radiation-induced liver disease in three-dimensional conformal radiation therapy for primary liver carcinoma: The risk factors and hepatic radiation tolerance. Int J Radiat Oncol Biol Phys 2006;65:

8 Reducing normal tissue injury d L. B. MARKS AND J. MA Gagliardi G, Lax I, Rutqvist LE, et al. Partial irradiation of the heart. Semin Radiat Oncol 2001;11: Marks LB, Kocak Z, Zhou S, et al. The association between the mean heart dose, mean lung dose, tumor location and RTassociated heart and lung toxicity [Abstract]. Proceedings of the American Society for Therapeutic Radiology and Oncology 47th Annual Meeting. Int J Radiat Oncol Biol Phys 2005; 63(Suppl. 1):S Forster KM, Smythe WR, Starkschall G, et al. Intensity-modulated radiotherapy following extrapleural pneumonectomy for the treatment of malignant mesothelioma: Clinical implementation. Int J Radiat Oncol Biol Phys 2003;55: Ahamad A, Stevens CW, Smythe WR, et al. Intensity-modulated radiation therapy: A novel approach to the management of malignant pleural mesothelioma. Int J Radiat Oncol Biol Phys 2002;55: Levegrun S, Hof H, Essig M, et al. Radiation-induced changes of brain tissue after radiosurgery in patients with arteriovenous malformations: Correlation with dose distribution parameters. Int J Radiat Oncol Biol Phys 2004;59: Flickinger JC, Kondziolka D, Pollock BE, et al. Complications from arteriovenous malformation radiosurgery: Multivariate analysis and risk modeling. Int J Radiat Oncol Biol Phys 1997;38: Fan M, Marks LB, Hollis D, et al. Can we predict radiationinduced changes in pulmonary function based on the sum of predicted regional dysfunction? J Clin Oncol 2001;19: Eisbruch A, Ten Haken RK, Kim HM, et al. Dose, volume, and function relationships in parotid salivary glands following conformal and intensity-modulated irradiation of head and neck cancer. Int J Radiat Oncol Biol Phys 1999;45: Goitein M, Abrams M, Rowell D, et al. Multi-dimensional treatment planning: II. Beam s eye-view, back projection, and projection through CT sections. Int J Radiat Oncol Biol Phys 1983;9: Chen GT, Spelbring DR, Pelizzari CA, et al. The use of beam s eye view volumetrics in the selection of non-coplanar radiation portals. Int J Radiat Oncol Biol Phys 1992;23: Verhey LJ. Comparison of three-dimensional conformal radiation therapy and intensity-modulated radiation therapy systems. Semin Radiat Oncol 1999;9: Webb S. The physical basis of IMRT and inverse planning. Br J Radiol 2003;76: Palta JR, Mackie TR. Intensity-modulated radiation therapy: The state of the art: Proceedings of 2003 summer school on IMRT. Madison, WI: AAPM/Medical Physics Publishing; Bortfeld T, Schmidt-Ullrich R, De Neve W, et al. Image-guided IMRT. Berlin: Springer; Nutting C, Dearnaley DP, Webb S. Intensity modulated radiation therapy: A clinical review. Br J Radiol 2000;73: Ezzell GA, Galvin JM, Low D, et al. Guidance document on delivery, treatment planning, and clinical implementation of IMRT: Report of the IMRT Subcommittee of the AAPM Radiation Therapy Committee. Med Phys 2003;30: Brahme A. Optimization of stationary and moving beam radiation therapy techniques. Radiother Oncol 1988;12: Radiation Therapy Collaborative Working Group. Intensitymodulated radiotherapy: Current status and issues of interest. Intensity modulated. Int J Radiat Oncol Biol Phys 2001;51: Bortfeld T. IMRT: A review and preview. Phys Med Biol 2006; 51:R363 R Glatstein E. Intensity-modulated radiation therapy: The inverse, the converse, and the perverse. Semin Radiat Oncol 2002;12: Newton KA, Spittle MF. An analysis of 40 cases treated by total thoracic irradiation. Clin Radiol 1969;20: Marks LB. The pulmonary effects of thoracic irradiation. Oncology (Will Park) 1994;8:89 106; discussion 100, Rosenzweig K, Gupta V, Jackson A, et al. Intensity-modulated radiation therapy for inoperable stage III non-small cell lung cancer [Abstract]. Proceedings of the 88th Annual Meeting of the American Radium Society. Oncology 2006;20(Suppl. 3): Yom SS, Liao Z, Liu HH, et al. Initial evaluation of treatmentrelated pneumonitis in advanced-stage non-small-cell lung cancer patients treated with concurrent chemotherapy and intensity-modulated radiotherapy. Int J Radiat Oncol Biol Phys 2007;68: Niemierko A, Goitein M. Modeling of normal tissue response to radiation: The critical volume model. Int J Radiat Oncol Biol Phys 1993;25: Bijl HP, van Luijk P, Coppes RP, et al. Influence of adjacent low-dose fields on tolerance to high doses of protons in rat cervical spinal cord. Int J Radiat Oncol Biol Phys 2006;64: Bortfeld T. Optimized planning using physical objectives and constraints. Semin Radiat Oncol 1999;9: Penagaricano JA, Shi C, Ratanatharathorn V. Evaluation of integral dose in cranio-spinal axis (CSA) irradiation with conventional and helical delivery. Technol Cancer Res Treat 2005;4: Chapet O, Fraass BA, Ten Haken RK. Multiple fields may offer better esophagus sparing without increased probability of lung toxicity in optimized IMRT of lung tumors. Int J Radiat Oncol Biol Phys 2006;65: Aoyama H, Westerly DC, Mackie TR, et al. Integral radiation dose to normal structures with conformal external beam radiation. Int J Radiat Oncol Biol Phys 2006;64: Shafman T, Marks LM, Das S, et al. Conservation of integraldose within, and dose-gradient between, shells surrounding targets: Physical realities and clinical consequences [Abstract]. ASTRO 44th Annual Meeting. Int J Radiat Oncol Biol Phys 2002;54(Suppl. 1): Fogarty GB, Hornby C, Ferguson HM, et al. Quality assurance in a radiation oncology unit: The chart round experience. Australas Radiol 2001;45: Kutcher GJ, Coia L, Gillin M, et al. Comprehensive QA for radiation oncology: Report of AAPM Radiation Therapy Committee Task Group 40. Med Phys 1994;21: Low DA. Quality assurance of intensity-modulated radiotherapy. Semin Radiat Oncol 2002;12: Bentel GC. Radiation therapy planning. 2nd ed. New York: McGraw-Hill; Fine LG, Kurtz I, Woolf AS, et al. Pathophysiology and nephron adaptation in chronic renal failure. In: Schrier RW, Gottschalk CW, editors. Disease of the kidney. 5th ed. Boston, MA: Little, Brown; p Maquet PGJ. Biomechanics of the hip. Berlin: Springer Verlag; Parke WW. The anatomy of the hip. In: Balderston RA, Rothman RH, Booth RE, et al., editors. The hip. Philadelphia: Lea Febiger; p Yorke ED, Jackson A, Rosenzweig KE, et al. Correlation of dosimetric factors and radiation pneumonitis for non-small-cell lung cancer patients in a recently completed dose escalation study. Int J Radiat Oncol Biol Phys 2005;63: Hope AJ, Lindsay PE, El Naqa I, et al. Modeling radiation pneumonitis risk with clinical, dosimetric, and spatial parameters. Int J Radiat Oncol Biol Phys 2006;65: Konings AW, Cotteleer F, Faber H, et al. Volume effects and region-dependent radiosensitivity of the parotid gland. Int J Radiat Oncol Biol Phys 2005;62: Eisbruch A, Kim HM, Terrell JE, et al. Xerostomia and its predictors following parotid-sparing irradiation of head-and-neck cancer. Int J Radiat Oncol Biol Phys 2001;50:

9 12 I. J. Radiation Oncology d Biology d Physics Volume 69, Number 1, Eisbruch A, Schwartz M, Rasch C, et al. Dysphagia and aspiration after chemoradiotherapy for head-and-neck cancer: Which anatomic structures are affected and can they be spared by IMRT? Int J Radiat Oncol Biol Phys 2004;60: Lu Y, Spelbring DR, Chen GTY. Functional dose-volume histograms for functionally heterogeneous normal organs. Phys Med Biol 1997;42: McGuire SM, Zhou S, Marks LB, et al. A methodology for using SPECT to reduce intensity-modulated radiation therapy (IMRT) dose to functioning lung. Int J Radiat Oncol Biol Phys 2006;66: Cheng CW, Das IJ. Treatment plan evaluation using dosevolume histogram (DVH) and spatial dose-volume histogram (zdvh). Int J Radiat Oncol Biol Phys 1999;43: Chao KS, Blanco AI, Dempsey JF. A conceptual model integrating spatial information to assess target volume coverage for IMRT treatment planning. Int J Radiat Oncol Biol Phys 2003;56: Kahn D, Zhou S, Ahn SJ, et al. Anatomically-correct dosimetric parameters may be better predictors for esophageal toxicity than are traditional CT-based metrics. Int J Radiat Oncol Biol Phys 2005;62: Poltinnikov IM, Fallon K, Xiao Y, et al. Combination of longitudinal and circumferential three-dimensional esophageal dose distribution predicts acute esophagitis in hypofractionated reirradiation of patients with non-small-cell lung cancer treated in stereotactic body frame. Int J Radiat Oncol Biol Phys 2005; 62: Novakova-Jiresova A, van Luijk P, van Goor H, et al. Pulmonary radiation injury: Identification of risk factors associated with regional hypersensitivity. Cancer Res 2005;65: van Luijk P, Novakova-Jiresova A, Faber H, et al. Radiation damage to the heart enhances early radiation-induced lung function loss. Cancer Res 2005;65: ICRU Report 50. Prescribing, recording, and reporting photon beam therapy. Bethesda, MD: International Commission on Radiation Units and Measurements; ICRU Report 62. Prescribing, recording, and reporting photon beam therapy (supplement to ICRU report 50). Bethesda, MD: International Commission on Radiation Units and Measurements; Ling CC, Humm J, Larson S, et al. Towards multidimensional radiotherapy (MD-CRT): Biological imaging and biological conformality. Int J Radiat Oncol Biol Phys 2000;47: Bradley J, Thorstad WL, Mutic S, et al. Impact of FDG-PET on radiation therapy volume delineation in non-small-cell lung cancer. Int J Radiat Oncol Biol Phys 2004;59: Ashamalla H, Rafla S, Parikh K, et al. The contribution of integrated PET/CT to the evolving definition of treatment volumes in radiation treatment planning in lung cancer. Int J Radiat Oncol Biol Phys 2005;63: Rietzel E, Liu AK, Doppke KP, et al. Design of 4D treatment planning target volumes. Int J Radiat Oncol Biol Phys 2006; 66: Herman MG. Clinical use of electronic portal imaging. Semin Radiat Oncol 2005;15: Webb S, Evans PM. Innovative techniques in radiation therapy: Editorial, overview, and crystal ball gaze to the future. Semin Radiat Oncol 2006;16: Yan D, Vicini F, Wong J, et al. Adaptive radiation therapy. Phys Med Biol 1997;42: Court LE, Tishler RB, Petit J, et al. Automatic online adaptive radiation therapy techniques for targets with significant shape change: A feasibility study. Phys Med Biol 2006;51: Chen J, Morin O, Aubin M, et al. Dose-guided radiation therapy with megavoltage cone-beam CT. Br J Radiol 2006;79: S87 S Letourneau D, Martinez AA, Lockman D, et al. Assessment of residual error for online cone-beam CT-guided treatment of prostate cancer patients. Int J Radiat Oncol Biol Phys 2005; 62: Pfeffer MR, Rabin T, Tsvang L, et al. Orbital lymphoma: Is it necessary to treat the entire orbit? Int J Radiat Oncol Biol Phys 2004;60: Senan S, van Sornsen de Koste J, Samson M, et al. Evaluation of a target contouring protocol for 3D conformal radiotherapy in non-small cell lung cancer. Radiother Oncol 1999;53: Rasch C, Steenbakkers R, van Herk M. Target definition in prostate, head, and neck. Semin Radiat Oncol 2005;15: