Stereotactic Body Radiotherapy for Primary Management of Early-Stage, Low- to Intermediate-Risk Prostate Cancer: Report of the American Society for Therapeutic Radiology and Oncology Emerging Technology Committee

Target localization and tracking 

An overriding principle of radiotherapy is to maximize the dose delivered to the tumor while sparing normal tissue to the greatest extent possible. This ideally increases the probability of tumor control and decreases the probability of normal tissue complication. In the case of the hypofractionation related to SBRT, this principle becomes even more important because any inaccuracy in patient setup can potentially result in severely harmful consequences.

Although target stereotaxis, immobilization, and tracking are essential to fully benefit from hypofractionation, a discussion of the numerous devices and techniques developed to addresses these needs is beyond the scope of this evaluation and will be the subject of subsequent reports.

Treatment delivery 

Typically, multiple, nonopposing, and often noncoplanar arcs, spread in a large solid angle with fairly equal weighting, are used to provide a high level of conformality (1). This method reduces the entrance dose and the volume of the irradiated normal tissue.

In deciding the number of beam directions and the relative beam weights, entrance dose should be kept to a modest level to prevent potential severe skin toxicity while keeping a uniform isotropic dose falloff (2). In the beam's-eye view, each beam is evaluated to coincide with the planning target volume (PTV). Typically, an isodose line of 60–80% provides 95% coverage of the PTV and is used to prescribe the treatment.

Three major criteria are used to evaluate the dosimetric properties of the SBRT plan: conformity index, high-dose spillage, and intermediate-dose spillage. The conformity index is defined as the ratio of the volume of the isodose shell that provides 95% coverage to the PTV. It is generally recommended that this ratio be kept to less than 1.2. High-dose spillage is defined as any volume receiving greater than 105% of the prescription and should be confined to the PTV. Intermediate-dose spillage is responsible for most of the toxicity associated with SBRT and can be characterized in two ways: (1) as the ratio of the 50% isodose volume to the PTV, or (2) by the dose to any point 2 cm away from the PTV. Either or both methods can be used, and intermediate-dose spillage should be minimized to limit toxicity. Dose–volume histogram criteria, commonly used with 3D-conformal radiotherapy or intensity-modulated radiotherapy (IMRT), have not been established.

In delivering SBRT, many commercially available systems are used. Sophisticated image guidance is a feature common to these treatment systems. Systems equipped with image guidance minimize the uncertainty associated with tumor localization. Most delivery systems also allow integration of immobilization devices. The most prevalent treatment systems used are the Novalis (BrainLAB, Feldkirchen, Germany), the TomoTherapy Hi-Art System (TomoTherapy, Madison, WI), the Varian Trilogy (Varian Medical Systems, Palo Alto, CA), the Elekta Synergy (Elekta, Norcross, GA), the Siemens Oncor and Artiste (Siemens Medical Solutions, Malvern, PA), and the CyberKnife Robotic Radiosurgery System (Accuray, Sunnyvale, CA).

The Novalis system uses a 6-MV linear accelerator with micro-multileaf collimators ranging in leaf thickness from 3 to 5.5 mm. Two KVp orthogonal X-ray cameras are mounted on the system to track bony landmarks or implanted fiducials in relation to the digitally reconstructed radiographs generated from computed tomography (CT) simulation. The patient is then aligned in the treatment position in accordance with identified positions of markers.

The TomoTherapy system uses a megavoltage CT to continuously image the patient, with the table continuously moving when IMRT is delivered throughout a full range of 360° rotations using a binary multileaf collimator system.

The Varian Trilogy and Elekta Synergy both use a cone-beam CT to provide real-time image guidance for repositioning. Elekta has recently acquired a company (3D Line Medical Systems, Norcross, GA) whose product was a full stereotactic system for cranial and extracranial radiotherapy treatments. The system is composed of a micro-multileaf collimator system with leaf thicknesses of 3 mm, 5 mm, or 7 mm, head and body frames for positioning and localization, dynamic patient support assembly with all translational, rotational, pitch, yaw, and roll movement, and an integrated optical tracking system for positioning. An inverse treatment-planning package with intensity-modulated arc therapy capability is also included.

The Siemens system is different in that the CT unit is linked to the accelerator via a shared tabletop, and it travels along rails. Once CT imaging is completed the tabletop rotates to the linear accelerator for treatment delivery, thereby providing a near real-time localization for treatment delivery.

The CyberKnife uses a frameless image-guided process to direct a robotic arm equipped with a linear accelerator. Translational and rotational movements of this robotic treatment are much broader and impossible to match with existing designs of linear accelerators. Two orthogonal diagnostic X-ray cameras are mounted on the ceiling to provide real-time imaging for tracking. Implanted fiducials or reliable bony landmarks are used to localize the tumor in real time for treatment delivery.

Rationale for prostate SBRT 

Prostate cancer lends itself to the use of SBRT because approximately 90% of cases present with disease clinically localized in the prostate. Furthermore, biologic models have suggested that prostate cancer may benefit from the use of fewer, larger treatment fractions, otherwise known as hypofractionation, which can vary in intensity. Stereotactic body radiotherapy, an accelerated form of hypofractionation, may be well suited for prostate cancer because it uses five or fewer fractions. Understanding the complexity of this issue requires a discussion of the radiobiology principals used.

Radiobiology (early vs. late effects and the α/β ratio) 

Conventional prostate radiotherapy is delivered in fraction sizes of 1.8–2.0 Gy. This method of fractionation emerged from the observation that late complications of radiotherapy have been reduced without an apparent compromise in local control. This approach is supported by the radiobiologic nature of surrounding normal tissues, such as the rectum.

Hypofractionation and SBRT present several potential advantages. Ideally, tumor control may be increased for a given level of late complications. Conversely, late complications may be reduced for a given level of tumor control. Patient convenience would increase with fewer fractions compared with standard external-beam radiotherapy treatment courses that extend for 7–9 weeks. Equipment utilization and staffing may be more efficient, which may translate to increased cost-efficiency.

Prostate cancer may not be typical of other tumors. The sensitivity of tissue to fractionation can be expressed as the α/β ratio. Tissues with a small α/β ratio (i.e., 2–4 Gy) are more sensitive to changes in fractionation than tissue with a large α/β ratio (i.e., >8 Gy). Tissues with a low α/β ratio are commonly referred to as “late-responding” because sequelae of treatment are generally seen years after treatment. Increasing the number of fractions generally spares late-responding tissues. Tumors are generally less sensitive to the effects of fractionation due to relatively large α/β ratios more characteristic of “early-responding” tissues. In general, increasing the number of fractions spares “late-responding” normal tissue while impacting minimally on the tumor. However, prostate cancer may not be an early-responding tumor. Instead it may be a late-responding tumor for which increasing fractionation may provide a sparing effect that limits or reduces the therapeutic ratio. If this is true, hypofractionation may be more effective for cell killing.

Alpha/beta ratio for prostate cancer 

Brenner and Hall (3) hypothesized that the α/β ratio for prostate cancer may be small because prostate tumors contain unusually small proportions of cycling cells (4). They estimated the α/β ratio for prostate cancer to be 1.5 Gy (95% confidence interval [CI], 0.8–2.2 Gy) by using clinical data to assume the linear and quadratic components of cell killing. Low-dose-rate brachytherapy results (5), used to estimate the linear (α, dose protraction-independent) component, together with external-beam radiotherapy (6) data, were used to derive an estimate of the α/β ratio for prostate cancer. Several investigators have repeated this method 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21. Using this method, most studies support this hypothesis and suggest that the α/β ratio is low, probably 1–4 Gy 22, 23. If the α/β ratio is low for prostate cancer this would support SBRT at the risk of increased late side effects. But if the α/β ratio is not low, SBRT may increase the risk for late toxicity and impair tumor control, thereby inverting the therapeutic ratio.

Some of the problems with the estimates of the α/β ratio include nonuniform dose distributions, tumor heterogeneity, varying relative biologic effectiveness (15), varying overall treatment times, and heterogeneity in tumor hypoxia. The linear-quadratic model is a simple application that does not take into account other fractionation-related phenomena, such as reoxygenation, redistribution, and repopulation. A detailed discussion of these issues has been published by Dasu (24).

King and Mayo (8) investigated a heterogeneity model and challenged that the α/β ratio is closer to 5 Gy, but subsequent studies have identified problems with heterogeneity modeling that have lowered this value (12). Dale and Jones (15) have indicated that if allowances are made for the increased relative biologic effectiveness of 125I and 103Pd, then the α/β ratio may be 1.0 Gy. With respect to estimates derived from interinstitutional and brachytherapy vs. external-beam radiotherapy comparisons, the best addressing data are from Brenner et al. (25), for which no interinstitutional comparisons and no comparisons of brachytherapy and external-beam radiotherapy were made. This yielded a value for α/β of 1.2 Gy (95% CI, 0.03–4.1 Gy). A similarly low α/β ratio of 1.5 Gy (95% CI, 1.2–1.8 Gy) has been supported by larger studies (10).

Fowler et al. (26) have modeled various hypofractionation regimens using the linear-quadratic model with the assumptions that the α/β ratio for prostate tumors is in the range of 1 to 2 Gy, and warn against the use of too few fractions (fewer than five) because this may limit the possibility of reoxygenation or redistribution of tumor cells into more sensitive phases of the cell cycle.

Alpha/beta ratio of the rectum 

The α/β ratio of the rectum is as important as the prostate α/β ratio in understanding what hypofractionation regimens will be beneficial or detrimental. The α/β ratio for the rectum is not known precisely. The generic value used for late-responding tissue such as the rectum is 3 Gy. In rodents, analyses of different experiments for late rectal damage by Brenner et al. (27) yielded 4.6 Gy (95% CI, 4.0–5.5 Gy). Van der Kogel et al. (28) reported 4.1 Gy (95% CI, 1.5–7.7 Gy), and Dewit et al. (29) found 4.4 Gy (95% CI, 1.6–7.7 Gy). Terry and Denekamp (30) reported a range of 3.1–5.1 Gy, whereas Dubray and Thames (31) found a range of 2.7 Gy (95% CI, 0.9–4.8 Gy) to 6.7 Gy (95% CI, 2.2–11.7 Gy). Gasinska et al. (32) found the α/β ratio to be 6.4 and 6.9 Gy for two different late rectal endpoints in mice. In summary, animal experiments suggest an α/β ratio for the rectum of 4–6 Gy.

Brenner (23) estimated the α/β ratio for late rectal bleeding using clinical results of hypofractionation with 1.8 Gy, 2 Gy, and 3 Gy per fraction data using the standard linear-quadratic model. The incidence of Radiation Therapy Oncology Group (RTOG) Grade 2 late rectal toxicity was fitted to the linear-quadratic model as a function of equivalent total dose if delivered in 2-Gy fractions using data points from rectal toxicity results from Memorial-Sloan Kettering Cancer Center (33), RTOG protocol 94-06 (34), the M. D. Anderson Cancer Center (35), and Akimoto et al. (36). Using this method the α/β ratio for the rectum was determined to be 5.4 ± 1.5. This is consistent with most estimates in animals.

If the α/β ratio for rectal damage is higher than that for prostate, then larger hypofractionated doses could be given with correspondingly larger clinical gains for the same constant late complication rates (26). If, however, the α/β ratio of late rectal reactions were smaller than that of the prostate, the incidence of complications would be increased. The relative increase is estimated to rise by factors of 1.15 at 15 fractions or 1.25 at 5 fractions (26). Such low numbers of fractions is the “worst case” likely, meaning that if a given complication were normally 5% it could rise to 6.3%, when using only 5 fractions (26).

Randomized clinical trials of hypofractionation with conventional radiotherapy, 3D conformal radiotherapy, and IMRT 

Clinical trials of hypofractionation are ongoing. Most trials have studied modest increases in daily fraction size while concerns of increased rectal toxicity predominate. Some investigators have introduced much more substantial hypofractionated regimens.

Nonrandomized clinical trials of hypofractionation with conventional radiotherapy, 3D conformal radiotherapy, and IMRT 

Clinical reports of SBRT 

Reports of SBRT have been limited. With the exception of Virginia Mason Medical Center and Stanford University, the majority of experiences are available in abstract form only.

Abstracts 

Prediction of social implications 

Prostate cancer is the leading cancer in men, with 186,320 new cases (25% of all cancer in men) estimated in 2008 by the American Cancer Society. Currently there are several standard local treatment options that include radical prostatectomy, prostate brachytherapy (seed implantation or high-dose-rate interstitial regimens), external-beam prostate irradiation alone (from a variety of radiation sources), and combinations of external-beam radiation and brachytherapy as a boost.

Prostate SBRT may provide an alternative that significantly reduces the overall cost of treatment and greatly reduces overall treatment time, which is favored by patients. Shorter treatment regimens may have significant impact on patients' ability to continue working, with minimal impact on productivity and resources, as long as acute toxicities and recovery times are not extended. Preliminary data suggest this is not the case, but additional studies are required to verify these preliminary observations. Longer follow-up is required to confirm clinical outcomes and quality-of-life measures.

As various SBRT regimens evolve, it is important to document via adequately powered clinical trials their relative efficacy and toxicity. This is particularly important in the setting of a shift in paradigm based on biologic modeling, to ensure that outcomes are at least equivalent to the long-established conventional fractionation regimens. Currently, evidence does not exist to establish that the hypofractionated techniques, especially with accelerated SBRT regimens, are equivalent to standard-fractionation radiation treatments. Prostate SBRT regimens are largely unverified in any venue other than relatively small single-institution trials with generally short follow-up.

Analysis of potential clinical issues 

The American Society for Therapeutic Radiology and Oncology maintains the position that new technologies and modifications of existing technologies representing significant paradigmatic shifts in treatment approach should be implemented into clinical practice in such a manner as to ensure safety, efficacy and, ideally, cost-effectiveness. Current conventionally fractionated courses of radiotherapy for prostate cancer at escalated doses are among the longest treatment courses for any disease, creating a major impetus to find effective shortened regimens.

The potential biologic advantages of hypofractionation in general are based on modeling of the linear-quadratic formula for the α/β ratio for prostate cancer cells and normal tissues, such as bladder and rectum. This model may not adequately address the complexities of tumor and normal tissue response, and therefore may not accurately be able to predict the dose per fraction that may be safely administered.

If the linear-quadratic equation is appropriate, the α/β ratio for prostate cancer may not be as low as some studies have suggested, when hypoxia and other factors are considered. Furthermore, it may not be the case that all risk groups will benefit equally; in fact, the intermediate-risk group has shown the most benefit from dose escalation in randomized trials, with less benefit to high- or low-risk groups. If biologic differences account for such variable outcomes, the same may be true for hypofractionation. Androgen deprivation therapy has also been suggested to influence the α/β estimate, and the interactions between androgen ablation and hypofractionation have also not been adequately addressed. To date, a variety of hypofractionation regimens have been used, and the optimal schema remains undefined.

Rectal toxicity is a major concern regarding prostate SBRT. Central to the success of dose escalation for prostate cancer with 3D conformal radiotherapy and IMRT has been the use of dose–volume constraints derived from thousands of patients treated with conventionally fractionated radiotherapy. Rectal dose constraints are not defined for fractions sizes used with SBRT. The late effects of even small volumes of rectum treated to doses in the range of 3–10 Gy are not known. A smaller, tighter margin on the prostate provides rectal sparing, but optimization of immobilization and precise daily real-time organ localization is required. A smaller margin also has the potential to result in poorer tumor control. Parameters for optimal dose per fraction, normal tissue dose constraints, PTV definitions, and dose distribution standards remain to be established for SBRT.

Technologic advances such as SBRT combined with optimum immobilization and organ localization may allow refinements in dose delivery precision to achieve the goal of minimal margins around the target structure while permitting dose acceleration. Clinical implementation of this technique will require a consistent investment in new technologies capable of achieving this precision, or poorer local control rates will probably result. These techniques are typically more time-consuming for the radiation oncologist and staff, involving the need for extensive contouring and closer oversight on treatment by the physician, as well as daily localization procedures on the part of the therapists. This increased time commitment should be offset by the fewer number of fractions used. The authors of this report believe further clinical trials addressing the uncertainties in the clinical implementation of this new approach to prostate cancer treatment should be conducted. The technique holds sufficient promise to warrant further investigation. If proven efficacious and safe, conventional and/or accelerated hypofractionation may provide social and economic benefits to prostate cancer patients as well.