Stereotactic radiotherapy is the logical step forward in the continual quest for improved dose distribution in radiation oncology. Advances in radiation oncology, often considered monumental, have all in principle simply been improvements in dosimetry. The advance from orthvoltage to megavoltage, and Cobalt teletherapy to linear accelerators, and from fluoroscopy to computerized tomography based treatments were all considered dangerous and/or frivolous until they became standard. Recent advances in imaging, computer hardware and software, now make extremely difficult targeting very possible. Many technologies were simultaneously developed for delivering highly targeted high dose treatments even before physicians clearly understood the potential of the engineering progress. New advances in technology continue to expand faster than the clinical application.
Along with advances in surgery and radiation oncology, have been some real benefits to cancer patients. The primary tumor is now rarely the dominant cause of lethality for patients with cancer. Local control rates for most common tumors are now very high. Complications of metastastatic disease remain the most common cause of death from cancer. Radiation and surgery are only rarely used for metastatic disease, and when they are, they are rarely used with curative intent. Chemotherapy remains the standard of care for metastatic disease, and cures are not expected with that single modality therapy.
The future utility of stereotactic radiation may not be limited to the simple improvement of local control for primary disease, or for palliation of solitary metastases. The opportunity now exists to combine stereotactic radiation with standard chemotherapy in the hope of yielding curative therapy for individuals with early metastastic presentation. The challenge of designing clinical studies that feature appropriate combinations of standard treatment and allow for use of continually changing technologies for stereotactic radiotherapy were discussed at the symposium that accompanied the presentations summarized in this volume of Acta Oncologica. In this editorial I will try and state the consensus.
Protocols composition
To date most stereotactic radiotherapy protocols have been single or limited institutional studies. A few cooperative oncology group studies have been opened but none are yet complete and none are randomized. Examples of existing protocols include RTOG 0235 and 0245 aimed at primary lung cancers and colorectal liver metatastases. Both of these studies were at least 5 years in development before they finally opened nationally in the United States. Issues that slowed development are instructive as we try and develop more national and international clinical studies. The issues include: definition of the protocol objective and identification of appropriate patient populations; determining allowable technology and implementation of quality assurance; settling on a specific dose and fractionation; combinations with chemotherapy; and opportunities for translational research to better understand cancer biology and tumor anatomy.
Protocol objective and appropriate patient populations
The standard of care for primary disease commonly includes both radiation and surgery. These are sometimes supplemented with adjuvant chemotherapy. Accepting this limited role for radiation, the main utility of stereotactic radiotherapy is to augment or replace surgery for tumor control.
If we alternatively accept the potential of radiation to perform in a paradigm shift that includes consolidation of chemotherapy for metastatic disease, clinical trial design becomes much more complex. Among patients treated with metastasis the standard for evaluation of therapy is RECIST. This methodology defines progression in a comprehensive way that includes presence of new metastases or growth of existing metastases. Other endpoints commonly used are to evaluate treatment results the frequency of complete and partial responses. None of these methods have satisfactory utility for stereotactic radiotherapy, which is not necessarily expected to produce a complete response in order to permanently control a tumor, and which cannot be well described by a parameter that focuses on all systemic disease while ignoring specifically targeted disease.
Studies on metastasis will need to include several secondary measures that feature both systemic failure and local failure parameters. The primary objective can be local control, progression free survival, or overall survival. Local control will be very high based on all phase I and phase II studies in the literature. Thus with local control as the primary objective, a phase III study will be a very small study. Arguably a phase III with local control as an endpoint is unethical since the result is so certain. A phase II study, with a local control endpoint, done as a cooperative study, is logical if the group is testing the feasibility of multiple institutions successfully performing an advanced technology study. It provides no other substantial purpose.
Randomized studies of stereotactic radiotherapy with median progression free survival or overall survival as the primary endpoints, is a design consistent with standard chemotherapy trials. Designing a randomized trial without chemotherapy in combination with the radiation however is difficult since micro-systemic disease cannot be ignored. Randomized studies of median survival can be limited to single metastases but logically should include at least three to five lesions as a maximum. Individuals with fewer metastases, in a variety of clinical studies, appear to have prolonged survivals compared to individuals with more than five lesions. A major problem with including patients with single lesions is ethical. The standard of care for solitary metastasis usually features local therapy (surgery±radiation). Thus a study that includes solitary metastasis, especially one that includes only solitary lesions, is likely to have poor accrual even if it passes a review board. An alternative to measuring median survival is to measure long term survival. Here a phase II study is sufficient, since the goal is to produce prolonged disease free survival. Long term disease free survival is rare for patients with multiple metastases. A flattening of a progression free survival rates with associated overall survival indicates potential cures. A study to detect cures may not need to be excessively large, but would have a much delayed final analysis. Admittedly however, designing a study wherein the goal is to begin curing patients with metastasis would be a powerful but probably over-optimistic study design.
Tumor types that are likely to be benefited by radiation metastectomy are those that are known to have organ specific tropism for their metastases. For example, it is accepted that women with breast cancer metastasis limited to bone have a better prognosis than those with less well behaved metastatic disease. Likewise, patients with prostate cancer seem to invariably have bone metastases and only a minority develops metastases to other organs. Patients with liver metastases from colorectal cancer are known to benefit from surgical metastectomy, but only non-central lesions can be treated by open surgery, leaving the majority available for study. Other examples include sarcoma and head & neck cancer, wherein metastases are most commonly to lung. To date studies have focused on non-resectable lung cancers, which moving forward might not be the best disease to study with regard to testing stereotactic techniques.
Quality assurance and dose and fractionation
Institutions presently performing stereotactic radiation use a wide variety of commercial techniques, usually locally modified by the investigators. All institutions use different methods to confirm target position, and there is no standard. Finally, all the systems have their advantages and disadvantages, and most clinicians will not want to abandon their proven targeting techniques. This problem will not dissipate with time since newer and “better” systems are constantly in development: the sciences and technology will continue to move forward while a clinical study is in progress. Quality assurance as with dosimetry techniques must keep to the basics. The standardization of clinical descriptions of gross tumor volume (GTV), clinical tumor volume (CTV) and planning tumor volume (PTV) is standardized and basic. As a matter of discussion, the GTV and CTV are the same, and the PTV should include no more than three standard deviations of motion and imaging error. The CTV expansion if any is often very difficult to define since tumor infiltration is ethereal on imaging studies. As with surgical margins, only a few millimeters margin is logical. Thus the PTV is typically about 5 to 10 mm laterally and anterior-posterior while superior-inferior margins are usually 10 mm. If treatment volume is at a premium due to normal tissue constraints, PTV need not be more than two standard deviations of target mobility error. Some correction for breathing is needed. To date virtually all studies have achieved very similar results independent of the technology. Quiet breathing without respiratory gating is the only technique that is consistently inferior and therefore probably should be avoided. Dose in stereotactic radiation must have very steep fall off thus placing the 60 – 80% isodose on the edge of the PTV is critical. Field sizes that allow 95% of the dose at the PTV edge to treatment plans that normalize to less than the 50% dose risk very high normal tissue exposure. Achieving a homogeneous or minimum tumor dose is unlikely to be important.
Much has been made of the required radiation dose for control of primary tumors in the lung. Less work has been done for metastases. It is important to remember that almost all studies of stereotactic hypofractionated radiotherapy and radiosurgery have very high tumor control rates. These control rates are typically near or above 80%. At these high levels of control, one might accept thatsome of the remaining 10 – 20% failures are due to inadequate dose. In order to develop the least toxic therapy that can be used for patients with the highest number of tumor deposits, the lowest effective tumor dose should be combined with the lowest peripheral dose. As mentioned above, when tumor control rates above 80% are consistently seen nearly independent of dose, the results strongly suggest that all the doses are sufficient and that other factors are responsible for failures. The commonly sited causes of failure being hypoxia or intrinsic biological factors (testable in translational studies), but one most admit the limitations of imaging and targeting and the training effect of both physician and therapist, likely account for some failures. In this regard newer imaging methods hold great promise.
Chemotherapy
Chemotherapy is effective therapy for metastases. Tumor responses are commonly seen with chemotherapy but cure is not expected. Permanent tumor control can be seen with chemotherapy of adult solid tumors, but the first growing lesion after a course of chemotherapy is commonly the previously known masses rather than newly developed lesions. The implications of this observation are two fold. First, median progression free survival is likely to be increased by adding SBRT since the index lesion rarely grows. Second, and perhaps more importantly, chemotherapy may at least delay new metastases and optimistically might down-stage a patient with systemic metastastases to an oligometastatic state. In that situation the potential added value of stereotactic radiation could be monumental. Thus chemotherapy is advised in clinical studies of metastases and probably also for combinations with primary tumor protocols. Chemotherapy is only advisable however when there are good chemotherapy choices, known to have high response rates. The choice of chemotherapy in protocol design is critical. A poor choice will impact toxicity and accrual. Chemotherapy can be used concurrent or sequential though most studies will want to use them sequentially to avoid confusion when ascribing toxicity. Toxicity should be measured using standardized international systems so that we will be able to compare toxicity levels between studies. The National Cancer Institute has suggested the CTCAE v 3.0.
Opportunities for translational research
This discussion can be arbitrarily long, and the conference did not focus on this aspect of SBRT. The opportunities for translational research are tremendous. For example, measurements of important classical radiotherapy parameters are possible. Stereotactic radiation allows the calculation of full dose response cures to measure lung tolerance. Patients can be offered radiosensitizers that target tumor specifically, and which take advantage of the selective radiation of tumor and exclusion of normal tissue. Alternatively, agents are being developed that prevent the fibrovascular complications of radiation, which can be given following irradiation. Mitigating side effects of radiation would make future stereotactic treatments safer.
Finally discoveries of which biological factors predict which patients are at risk of developing metastases or which specific genetics dictate specific organ tropism might be possible. Understanding the genetic determinants of organ specificity would naturally lead to advancement of the current clinical staging systems and clinical trial design. One can expect staging systems that pay more attention to stages of metastatic disease (vis M0, M1 [oligo metastasis], M2 [single organ metastasis], M3 [widely disseminated metastasis]) to emerge from genetic array studies of clinical material. This will help better stratify patients when designing clinical studies, and potentially change the prognosis of patients with metastases.
Future
There is great hope that stereotactic radiation might reduce the need for invasive surgical procedures and provide consolidation of chemotherapy. Optimists see it as an agent that might finally cure some patients with metastatic disease. Cooperative clinical trials between multiple institutions will be a challenge. Design of these trials will require technology independent treatment rules since the technology being used when the results of a study are finally published will have already moved forward. Logically treatment of metastases has great potential since existing data is consistently excellent with regard to local control. A key will be to design studies that do not increase toxicity while still producing high control rates. Great effort must be paid to reduction of PTV to the minimum required, while also lowering the tumor dose to the minimum required. Designing studies that decrease treatment volume are in accordance with natural clinical instinct and should be easy to design and execute. Studies aimed at decreasing dose are important but may be harder to perform.
Studies featuring stereotactic radiation for both primary disease and for metastasis are among the most exciting developments in cancer care with a real chance to change the course of cancer survival. Trials in SBRT therefore, though difficult to develop, should be a very high priority for the clinical oncology community.