Stereotactic Body Radiation Therapy: Scope of the Literature

Stereotactic body radiation therapy (SBRT) is derived from the techniques of stereotactic radiosurgery used to treat lesions in the brain and spine. It combines multiple finely collimated radiation beams and stereotaxy to deliver a high dose of radiation to an extracranial target in the body in a single dose or a few fractions. This review provides a broad overview of the current state of SBRT for solid malignant tumors. Reviewers identified a total of 124 relevant studies. To our knowledge, no published comparative studies address the relative effectiveness and safety of SBRT versus other forms of external-beam radiation therapy. Stereotactic body radiation therapy seems to be widely diffused as a treatment of various types of cancer, although most studies have focused only on its use for treating thoracic tumors. Comparative studies are needed to provide evidence that the theoretical advantages of SBRT over other radiation therapies actually occur in the clinical setting; this area is currently being studied in only 1 small trial.

R adiation has been used for the treatment of cancer since 1896, shortly after Roentgen's discovery of x-rays. External-beam radiation therapy is the most common method of radiation delivery; brachytherapy (placement of radioactive materials in or adjacent to the tumor) is another common method. Before external-beam radiation therapy, the target lesion is defined on imaging studies; a simulation is constructed; and a treatment plan is generated by radiation oncologists, dosimetrists, and radiation physicists. The use of 3-dimensional imaging techniques, such as computed tomography, positron emission tomography, or magnetic resonance imaging, for treatment simulation and planning has improved the accuracy of external-beam radiation therapy.
The software used in planning treatment incorporates data from the 3-dimensional images to more precisely define the target, normal tissues, and the isocenter at which the radiation will be delivered most intensely. Images taken at different times can be incorporated into the planning algorithms to account for movement of the target and surrounding structures during treatment, such as with respiration-this is called 4-dimensional conformal radiation therapy (1).
Another advance in the delivery of radiation is improved shaping of the radiation beam delivered to the tumor by using multileaf collimator systems located in the linear accelerator (linac). The multileaf collimator consists of multiple metal "leaves" that can be adjusted individually to shape the radiation beam to the contour of the tumor. Methods for reducing movement of the patient on the treatment table and of the tumor during treatment have also improved the accuracy of radiation delivery. For treatment of areas in the body, immobilization is accomplished by using abdominal-compression devices, vacuum pillows, masks, and stereotactic frames that fit tightly around the patient and are fixed to the treatment table.
Some delivery devices can track the position of the tumor and surrounding structures during the treatment, using implanted fiducials (radio-opaque markers), skin markers, or bony structures as reference points. A more recent trend is the use of image-guided radiation therapy before (or during) a treatment session to assess changes in the movement or shape of the tumor and the accuracy of delivery in real time. Digital orthogonal radiography is commonly done at the time of treatment, and results are compared with computed tomography-generated digital radiographs.
The improved ability to shape the radiation beam and spare surrounding tissues has led to a greater interest in reducing the number of fractions over which treatment is delivered. Treatment with 3-dimensional conformal radiation therapy or intensity-modulated radiation therapy is typically divided into 25 to 50 fractions (approximately 2 Gy per fraction) delivered 5 days per week for approximately 5 to 10 weeks. Stereotactic radiosurgery, in contrast, delivers a high dose of radiation to cranial tumors in a single session. Stereotactic body radiation therapy is typically delivered in 1 to 5 fractions, with a typical total dose of 20 to 60 Gy. The Figure shows a device with SBRT capabilities.
As of September 2009, we had identified 384 facilities listed in the AHA Guide to the Health Care Field 2009 Edition (2) that specifically stated their capacity to provide SBRT. Lanni and colleagues (3) compared the hospital charges for technical and professional treatment with SBRT with those of intensity-modulated radiation therapy and standard fractionated radiation therapy for medically inoperable non-small cell lung cancer (NSCLC) on the basis of the average number of delivered fractions (3). The average charge was $55 705 for 35 fractions of 3dimensional conformal radiation therapy, $146 570 for 35 fractions of intensity-modulated radiation therapy, and $52 471 for 4 fractions of SBRT (3). The expected Medicare reimbursements (based on the 2010 Ambulatory Payment Classification) for the respective therapies were $13 639, $22 747, and $10 616 (3).
The Agency for Healthcare Research and Quality Effective Health Care Program requested a technical brief on SBRT. A technical brief is not intended to assess comparative effectiveness, but to provide an overview of key issues related to an emerging diagnostic or therapeutic technology. The goal of this technical brief is to provide a broad overview of the current state of SBRT for solid malignant tumors. Although some authors and organizations include treatment of lesions in the spine as a form of SBRT, others consider this stereotactic radiosurgery. Several terms have been used to describe SBRT in the published literature, such as stereotactic radiother-apy, fractionated stereotactic radiosurgery, hypofractionated stereotactic radiosurgery, and staged radiosurgery. For the purposes of this review, we will use the term SBRT and will not discuss treatment of spinal lesions. Other terms used in this article are defined in the Glossary.

Protocol
The protocol for the technical brief was developed in conjunction with the Agency for Healthcare Research and Quality. The protocol for the literature scan was developed by the ECRI Institute Evidence-based Practice Center, Plymouth Meeting, Pennsylvania, with input from ECRI Institute's medical physicists. Peer and public review of an early draft of the technical brief led to changes in terminology and scope based on recent consensus definitions. Proton, electron, and carbon ion therapy are outside the scope of this technical brief.

Search Strategy
We searched for published English-language studies by using the Ovid platform in MEDLINE; EMBASE; and Cochrane databases, including the Health Technology Assessment Database, from January 2000 to December 2010. Search terms included (but were not limited to) hypofractionated, stereotaxis, stereotactic, and single-fraction and text and Medical Subject Heading terms for cancer and devices. The full search strategy can be found in Appendix Tables 1 and 2 (available at www.annals.org).
We used Google to search for gray literature applicable to the background information, descriptions of instrumentation and accessories, and utilization. We separately searched Windhover Online, Healthcare News, The Gray Sheet, The Wall Street Journal, and Clinica. We visited the Web sites of related professional association and organizations (for example, the International RadioSurgery Association). We found information on instrumentation by searching manufacturers' Web sites and the U.S. Food and Drug Administration's Center for Devices and Radiological Health Web site (www.fda.gov /cdrh). Instrumentation information can be found in Supplement 1 (available at www.annals.org). We obtained additional information on device specifications and compatible accessories through interviews with manufacturers (available in the full report [4]).

Study Selection
Two reviewers screened abstracts and full-text articles, and a third reviewer assisted with disagreements. Eligible studies were clinical studies of any design; were published in English; had at least 3 patients with solid malignant tumors in the body (excluding head and spine) who received SBRT in 10 or fewer fractions; and reported any clinical outcomes, adverse events, or both. Studies not eligible for data extraction included those on treatment planning (for example, dosing) and treatment delivery (for example, accuracy) that did not report patient outcomes. A

Key Summary Points
Advances in planning and delivering radiation treatment have led to greater interest and capabilities to treat smaller and hard-to-target tumors while reducing treatment time.
Stereotactic body radiation therapy (SBRT) is typically delivered in 1 to 5 fractions, with a typical total dose of 20 to 60 Gy.
Stereotactic body radiation therapy seems to be widely used for treating various types of cancer, although most studies have focused only on its use to treat thoracic tumors. Fewer than 10 studies each for tumors of the pancreas, liver, colon, uterus, pelvis, sacrum, kidney, and prostate were found in this technical brief.
The American Association of Physicists in Medicine Task Group on SBRT has emphasized the importance of having well-trained and dedicated staff for providing SBRT in a safe environment.
No published comparative studies address the relative effectiveness and safety of SBRT versus other forms of external-beam radiation therapy. Only 1 small, ongoing trial is making such a comparison.
Comparative studies are needed to provide evidence that the theoretical advantages of SBRT over other radiotherapies actually occur in the clinical setting.
Future studies may help to determine the optimal number of radiation fractions, minimum and maximum doses per fraction, maximum number and diameter of lesions for various locations, and radiobiological explanations for the efficacy of SBRT.
detailed list of the excluded articles and primary reason for exclusion can be found in the full report (4).

Data Extraction
We created standardized data extraction forms, and 2 reviewers entered data into the SRS 4.0 database (Mobius Analytics, Ottawa, Ontario, Canada). The information extracted (if reported) included cancer type, inclusion or exclusion criteria, patient outcomes, adverse events or harms, study design characteristics, treatment characteristics, patient characteristics, instrumentation, algorithms, and quality assurance or training procedures.

Data Analysis and Synthesis
We performed descriptive statistics for patient characteristics and length of follow-up. We described study types, patient populations, prior or concurrent treatment, devices and algorithms used, outcomes assessed, patient selection criteria reported by the studies, and adverse events.

Role of the Funding Source
The ECRI Institute Evidence-based Practice Center prepared this report, with funding from the Agency for Healthcare Research and Quality. Staff from the Agency for Healthcare Research and Quality participated in the formulation of the research questions and reviewed the methods and the draft report but were not involved in study selection, data extraction, or drafting of the manuscript for publication.

RESULTS
Of the 5585 citations screened, 124 relevant studies were identified. These studies are summarized by cancer type in Appendix Table 3 (available at www.annals.org), and the study selection process is shown in the Appendix Figure (available at www.annals.org).

Tumor Sites Where SBRT Has Been Used
The tumor sites most commonly represented in the included studies were the lung or thorax (n ϭ 68). We found 27 studies of tumors in the pancreas, liver, and colon and fewer than 10 studies each of tumors in the uterus, pelvis, kidney, thyroid, and prostate. Ten studies included patients with various treatment sites.

Radiation, Instrumentation, and Algorithms Used
Photon radiation was used in all of the included studies. The instrumentations most frequently reported included modified linacs, CyberKnife (Accuray, Sunnyvale, California), Novalis Shaped Beam or Clinac (Varian Medical Systems, Palo Alto, California), TomoTherapy Hi-ART (TomoTherapy, Madison, Wisconsin), a fusion of Figure. A linear accelerator-based stereotactic radiosurgery and stereotactic body radiation therapy system.
The system incorporates a high-definition, 120-leaf multileaf collimator; robotic couch; cone-beam computed tomography; and stereoradiograph target-verification system. (Photo courtesy of Varian Medical Systems, Palo Alto, California. All rights reserved.)

Glossary
Brachytherapy: The placement of temporary or permanent radioactive material inside the body. Cone-beam computed tomography: Often built into linear accelerators and uses a cone-shaped radiation field to obtain images over 1 to several minutes. The image is then modified to allow viewing in a usual coronal, transaxial, and sagittal format to allow high-precision targeting. Computed tomography: Produces images based on a cross-sectional view of the body obtained from different projections in a given plane. Coplanar: Within the same plane. External-beam radiation therapy: Radiation therapy delivered from outside the body. Fiducials: Markers that help to precisely identify the targeted tumor location. They may be located on a surface or surgically implanted for tumor locations throughout the body. Four-dimensional conformal radiation therapy: Measures the natural tumor motion that occurs during treatment, such as from breathing. The measurements are included in the treatment plan to ensure that the tumor is targeted throughout the treatment course. Fraction: A prescribed total treatment dose divided into smaller amounts. Image-guided radiation therapy: The use of imaging methods (such as computed tomography) to assist in targeting a lesion during radiation treatment. Intensity-modulated radiation therapy: Therapy that delivers radiation beams with varying intensities. Isocenter: The center of the treatment area in relation to the paths of the radiation beams. Linear accelerator (linac): Radiography machine that emits photon-or electron-radiation megavolts redirected in many arcs for treating benign or malignant lesions throughout the body. Magnetic resonance imaging: Produces images by using high-field magnets and radiofrequency energies that can be acquired along any direction without using ionizing radiation. It is particularly useful for soft-tissue discrimination. Multileaf collimator: Located in the linear particle accelerator and consists of multiple metal "leaves" that can be adjusted individually to shape the radiation beam to the contour of the tumor and to variably modulate the dose. Noncoplanar: Within different planes. Positron emission tomography: Produces images to evaluate biological function by detecting gamma radiation emitted from a positron-emitting compound that is injected into the patient (fluorodeoxyglucose is common). Respiratory gating: Tracking patient's normal respiratory cycle with an infrared camera and a marker placed on the chest or abdomen or other methods. Stereotaxy: Three-dimensional target localization. Stereotactic frames provide reference points (coordinates) in 3 dimensions. Three-dimensional conformal radiation therapy: A 3-dimensional planning system to deliver radiation to the tumor.
computed tomography and linac, and the Synergy system (Elekta, Stockholm, Sweden).
Most studies described the treatment-planning techniques and treatment-delivery processes. Studies reported the use of 1 to 12 coplanar or noncoplanar radiation beams for treatment. Various body-immobilization techniques were used, including the Alpha Cradle (Smithers Medical, North Canton, Ohio) and the Stereotactic Body Frame (Elekta).
Computed tomography, magnetic resonance imaging, and positron emission tomography were often used to plan treatment. Treatment planning was conducted on software systems that typically were specific to the device used during treatment. Breath-holding, respiratory gating, and abdominal-compression techniques were used to control for respiratory movement. Finally, the type of image guidance (megavoltage or kilovoltage) used during treatment (that is, just before treatment began) included computed tomography, cone-beam computed tomography, and orthogonal radiography.
Doses and fractions of SBRT varied on the basis of such factors as the type of cancer and location of the tumor. Most published studies described delivery in 1 to 5 fractions; however, 14 studies delivered treatment in more than 5 fractions but still referred to this as SBRT (5-18).

Study Design, Study Size, and Follow-up
We identified 67 prospective single-group studies (Supplement 2, available at www.annals.org) and 57 retrospective single-group studies (Supplement 3, available at www.annals.org) of SBRT that met our inclusion criteria. The 2 categories of tumor sites with the most studies and patients were the lung or thorax, with 68 studies and 4418 patients, and gastrointestinal sites (colon, liver, or pancreas), with 27 studies and 1281 patients (Appendix Table  3). The shortest mean and median follow-ups were in the studies of multiple cancer sites (8.2 and 12.9 months, respectively; range, 1 to 95 months). Studies of tumors involving the pelvis, sacrum, and uterus had the longest mean and median follow-ups (31 and 33 months, respectively; range, 2 to 77 months).

Comparator Used in Comparative Studies
None of the included studies compared SBRT with another form of radiation treatment. We searched ClinicalTrials.gov and identified 50 ongoing SBRT trials. The trials include metastatic breast cancer and primary cancer of the biliary tract, kidney, liver, lung (principally NSCLC), pancreas, prostate, and unspecified sites. Only 1 of these ongoing trials involves a direct comparison of SBRT with a different form of radiation therapy. This trial (ClinicalTrials.gov identifier: NCT00870116) is a nonrandomized comparison of SBRT delivered by CyberKnife (enrollment target of 20 patients) versus SBRT delivered by linac (enrollment target of 80 patients) versus conformal radiation therapy without respiratory tracking (enrollment target of 20 patients) for NSCLC. The primary outcome measure is local control. The trial commenced in April 2009 in France and is still recruiting patients. Three other comparative trials are designed with historical controls, including 1 each for metastatic breast cancer (ClinicalTrials .gov identifier: NCT00167414), NSCLC (ClinicalTrials .gov identifier: NCT00727350), and pancreatic cancer (ClinicalTrials.gov identifier: NCT00350142). A randomized lung cancer trial based in the Netherlands (Clinical Trials.gov identifier: NCT00687986) is comparing SBRT with primary resection of the tumor. The primary outcomes are local control, regional control, quality of life, and treatment costs. The enrollment target was 960 patients, and completion was expected in December 2013; however, the trial was terminated in April 2011 because of poor recruitment. A trial conducted in China (Clinical Trials.gov identifier: NCT00840749) will compare SBRT with surgical resection in NSCLC. The enrollment target is 1030 patients, with planned completion also in 2013. Another trial (ClinicalTrials.gov identifier: NCT00843726), conducted at Roswell Park Cancer Institute, Buffalo, New York, will randomly assign 98 patients to either 1 or 3 fractions of SBRT for NSCLC. Appendix Table 4 (available at www.annals.org) includes further information on these 6 or 7 comparative trials. Additional ongoing trials of SBRT are described in the full report (4).

Patient Selection Criteria
The most common selection criteria were inoperable tumors or patients declining surgery. Most studies required biopsy-proven disease before treatment and a minimum level of performance on the Karnofsky Performance Scale or the World Health Organization/Eastern Cooperative Oncology Group Performance Status scale, typically excluding patients with severe illness or disability. Patients who had received prior radiation therapy were typically excluded unless a minimum interval had elapsed before SBRT. Fifty-eight studies reported prior or concurrent treatment (for example, surgery, chemotherapy, and radiofrequency ablation).

Outcomes Measured
The outcomes most frequently assessed were tumor response (n ϭ 55) (

Adverse Events Reported
Many of the studies used the Radiation Therapy Oncology Group criteria and Common Toxicity Criteria, Version 2.0, to grade acute and late toxicity. A total of 113 studies reported acute or late adverse events. The most frequently reported adverse events were pain, fatigue, nausea, bleeding, and diarrhea. Study authors noted that some patients with these events had received prior treatment for Review Stereotactic Body Radiation Therapy cancer, received SBRT for recurring cancer, or had comorbid conditions. One study of SBRT for renal cell cancer reported no adverse effects of treatment (54).

DISCUSSION
Stereotactic body radiation therapy is a treatment option that is likely to be attractive to patients because of the convenience of fewer treatment sessions. The radiation oncologist may decide whether to use SBRT in any given patient on the basis of the patient's radiation history (in particular, prior radiation of the site to be treated), proposed treatment volume, function of the involved or nearby organs, the patient's capacity for recovery, patient preference, the number of tumor sites, and many other cancer-related factors (116).
Although many published studies describe SBRT, several limitations in the available literature raise concerns about the widespread use of SBRT. To date, no published comparative studies (either randomized or nonrandomized prospective studies) have addressed the relative effectiveness and safety of SBRT versus other forms of externalbeam radiation therapy. Also, SBRT seems to be widely disseminated for treatment of various cancer types, although most of the studies (n ϭ 68) were for SBRT for tumors located in the lung or thorax. We found fewer than 10 studies each of tumors of the pancreas, liver, colon, uterus, pelvis, sacrum, kidney, and prostate. A Medicare Evidence Development and Coverage Advisory Committee proceeding from April 2010 considered the evidence for radiation therapy for localized prostate cancer (117). The panel of 15 experts was asked to state its level of confidence in the evidence for improvement of mortality (survival and death rates), functional outcomes (erectile dysfunction, urinary incontinence, and fecal incontinence), and adverse events (rectal fistula, radiation burns, and infection) with SBRT compared with classical fractionated external-beam radiation therapy for localized prostate cancer. On a scale of 1 (low confidence) to 5 (high confidence), the average score was 1.07 for mortality, 1.13 for functional outcomes, and 1.33 for adverse events (117).
To date, 7 systematic reviews assessing SBRT for cancer have been published. Three assessed its use in NSCLC (118 -120), 2 in multiple sites (for example, liver and pancreas) (121, 122), 1 in liver cancer (123), and 1 in pulmonary metastasis (124). The general consensus among these reviews is that although single-group studies show potentially promising results for various cancer sites, prospective studies are necessary to determine the efficacy of SBRT compared with other available treatment options (for example, surgery or radiation therapy). Appendix Table 5 (available at www.annals.org) provides more detail regarding these reviews.
Such groups as the American Association of Physicists in Medicine have also urged participation in trials sponsored by the National Cancer Institute or in trials run by the National Cancer Institute-sponsored Radiation Therapy Oncology Group, a multi-institutional research cooperative. In a recent guidance document, the American Association of Physicists in Medicine pointed out that protocol-driven treatment in the context of such studies would reflect the guidelines produced by experts in the field (125). Future studies may help to determine the optimal number of radiation fractions, minimum and maximum doses per fraction, maximum number and diameter of lesions for various locations, and radiobiological explanations for the efficacy of SBRT.
Chang and Timmerman (126) point out that more multicenter trials are needed to corroborate results in single-institution trials and, like others (116), noted that longer-term follow-up will be necessary to assess late toxicities from SBRT. In addition, they point out that although the relative effectiveness and safety of the various devices available for delivering SBRT need further study, the training and experience of the operators may be of greater importance (126). The American Association of Physicists in Medicine Task Group on SBRT has emphasized the importance of having well-trained and dedicated staff for providing SBRT and points out the procedures that need to be performed by such personnel to ensure patient safety (125).
In summary, there are many publications on SBRT for cancer, principally NSCLC. Comparative studies (preferably randomized trials, but at least trials with concurrent controls) are needed to provide convincing evidence that the theoretical advantages of SBRT over other radiation therapies actually occur in the clinical setting. Appropriate comparators for SBRT studies may depend on the treatment site, tumor size, or feasibility of other forms of radiation therapy (for example, intensity-modulated radiation therapy or proton-beam therapy) or surgery for the patient population. At present, only 1 small, ongoing trial is making such a comparison. Consequently, a full systematic review of the current literature cannot answer questions on the effectiveness and safety of SBRT compared with other radiotherapy interventions. One large ongoing trial scheduled for completion in 2013 has the potential to answer questions about the effectiveness and safety of SBRT compared with surgical resection in patients with resectable early-stage lung cancer.  quality of life; RC ϭ regional control; RT ϭ radiation therapy; SBRT ϭ stereotactic body radiation therapy; SRT ϭ stereotactic radiation therapy.