New Radiation Modalities

 

 

 

 

 

 

 

 

A Brief History

In the early days of radiation therapy, low energy beams were used, and were set up using external Landmarks. Over the first few decades, improvements came in increasing the energy of the treatment beams and in later instituting the use of fluoroscopic simulators. We went from x-ray beams generated in x-ray tubes with energies in the same range as current diagnostic machines up to megavoltage beams using cesium, cobalt, and van de Graaf generator. In the late 1960’s linear accelerators began to be commercially available, and now have become the norm for treatment in the United States. Modem linear accelerators typically have dual energies (at least in larger centers) with a lower energy of around 6 megavotts for more superficial targets and a higher energy of 15-20 megavolts for deeper targets. In general, higher beam energies will result in more sparing of organs at risk when treating deep-seated structures and tumors. Smaller centers may have only a single beam energy, and typically this is 6 megavolts. This becomes important usually only in very large patients, where doses to skin and critical structures may be somewhat higher. Although it is possible to make higher beam energies avail- able, we have reached the point where further increases in beam energy will not result in measurable sparing of organs at risk.

Concurrent with the development of higher treatment beam energies came the development of better methods of target localization. The biggest early advance was the invention of the fluoroscopic orthovoltage simulator. More recently this has become less important with the advent of CT scan simulation, although conventional simulation is still used for many situations, particularly palliative treatment of the brain and bone metastases. 3-D conformal radiation is now the standard for most deep-seated tumors.This employs the use of a diagnostic quality CT scan with the patient in the treatment position, including any immobilization devices. Volumes of tumor and critical structures are outlined, and a three-dimensional computer generated plan is derived. The assumption is that the tumor and critical structures are in the same place every day during treatment and that they do not move between diagnostic scanning and treatment. That assumption is rarely correct, hence many new developments. Practically all new developments are aimed at giving higher doses to tumor and lower doses to critical structures, coupled with better localization of areas at risk.

NEW MODALITIES:

IMRT —

IMRT stands for intensity modulated radiation therapy. Basically, this involves the use of one of two basic techniques to give differential doses to different areas of the treatment field. The idea is to try to spare normal tissue, thus either reducing complications and toxidty or allowing higher doses to areas at risk. There are two basic methods of achieving this result. The first uses a milled metal block that uses differential thicknesses to achieve differential doses at various areas of the field. This requires considerable physics support in order to verify dosimetry of the block, but is overall simpler than the second method of achieving variable doses. This method of producing IMRT has recently been approved for payment under a separate CPT code of its own. In some ways, it is less convenient for treatment because each treated field requires a separate compensator block that must be manually placed prior to treatment of the field.

The second means of achieving IMRT is through the use of multileaf collimators and variable coverage of the field by the vanes of the multileaf collimator. A multileaf collimator uses multiple metal “vanes” that can be driven across the treatment field to block the radiation beam. By varying where they cover and for how long, you can vary the dose to structures in the field. Although it is possible to do this in a dynamic fashion, most institutions use the simpler (and more verifiable) step-and-shoot technique, wherein the treatment of a field is divided into several segments and these are treated in a sequential fashion. This allows verification of each segment with film or other types of dosimetric measurements. In other words, there may be 5-7 (or more) fields (also called ports or portals), and each will have several segments, or different MLC settings.

With this advance has come a new problem. IMRT assumes that we know with a high degree of reliability what we need to treat and what we need to spare, and that those structures will always be where they were at the time the planning was done. We are better defining at-risk areas by studying data on failures. Anatomical areas at risk are being better defined by studies of old anatomic studies, use of higher quality imaging with high speed CT scanners and MRI, and the use of functional imaging with PET and soon with dextrose coated ferric materials to look at lymph nodes. Thus, we are getting a better handle on these issues. The other issue, knowing that things are where they are thought to be, is more difficult. Immobilization is useful for some sites, but is more problematic in areas such as the abdomen and pelvis. These issues are extremely important in IMRT because of the very nature of the procedure. By definition, the high dose volume is concentrated much more than with more traditional therapy. A small amount of movement of expected structures can result in completely missing the target, and in the worst case scenario actually put the high dose entirely in critical structures. Therefore, IMRT puts great requirements of quality control on the radiation oncology department In some cases, weight loss and other factors may require repeated scanning and planning. IMRT requires an enormous amount of physics support, and is more time consuming for the radiation oncology department, but also is reimbursed at a much higher rate than conventional radiation, and is widely touted by institutions because of the economic issues. Therefore, it is worthwhile to look at the data regarding efficacy and safety and see if the clinical data support the greatly increased costs.

 

Head and neck cancers —

This is probably the area where IMRT has most dearly been shown to have efficacy. The biggest advantage is the sparing of salivary tissue. If doses to the parotid glands are kept below 20-26 Gy, then salivary function tends to be preserved. This eliminates for many patients the most troublesome after effect of head and neck irradiation. A related and perhaps not predictable effect of the institution of IMRT treatment is that radiation oncologists have had to become more expert in the lymphatic anatomy and function in the head and neck area, and ultimately this may yield result improvements as great as those achieved with IMRT.

 

Prostate, other pelvic tumors —

The goal of IMRT in the pelvis is to allow treatment of nodal stations at risk with reduced doses to small bowel and to allow treatment of the prostate, rectum, or gynecologic organ with reduced doses to bladder and rectum (except, obviously, when the rectum is the target.) Again, this requires more anatomic knowledge on the part of the radiation oncologist. There appears to be little doubt that IMRT can in fact accomplish the stated aims. The real question is whether this will afford any real advantage to the patient, and that question remains unanswered. There is considerable data for the prostate, for instance, showing that prostate movement during treatment can be substantial enough to cause perturbations of the planned dose placement.

 

Liver, pancreas, kidney —

IMRT is still under investigation for these tumors. They present with two problems. The first is that there is organ motion with respiration. This makes treatment of very conformal plans very difficult and inherently dangerous without substantial controls of the entire treatment process. The second problem is that there are numerous sensitive organs at risk of damage in immediate proximity to the desired treatment area. Both of these are difficult to overcome. Organ motion difficulties can be partially overcome by simply adding larger margins, but there are limits to this approach and this also further exacerbates the problems of organs at risk since it can push high dose areas into critical structures. At the present time, it can fairly be said that IMRT for treatment of these organs is under investigation and may be promising if the difficulties are overcome.

 

Lung —

IMRT for lung tumors is widely agreed to be investigational and probably not appropriate for i.e in most community centers. Even well funded academic institutions are having difficulty with this area. The biggest problem is the motion of the tumor during treatment. This is exacerbated by dose heterogeneity that is much greater in the lung than in other places because of the air/tissue interfaces and the complex interactions of radiation at those interfaces. Until recently the National Cancer Institute prohibited the use of IMRT in lung protocols they were funding for cooperative groups. They have recently softened that stance only slightly, allowing IMRT in these protocols only with substantial controls for the problems.

 

Breast —

IMRT can reduce skin and lung radiation dose and virtually eliminate dose to the heart, and is therefore is recommended for treatment of left-sided breast cancers. It is not felt to be necessary for the right side, although many centers use a simple type of IMRT on the right side without extra charge to the patient.

IMRT has considerable promise, but also presents substantial challenges. Unfortunately, economic issues are clouding our vision. When CMS approved the CPT and APC codes for IMRT, they did so with the understanding that some 25-30% of patients would be treated with IMRT, and it is not appropriate to talk about treating a very large fraction of patients with this technology. There are also many substantive issues that must be answered by future follow-up of large numbers of treated patients, such as what the late toxicities will be and what risk there is of second malignancies when large volumes of tissue are treated to low radiation doses.

IGRT —

IGRT stands for image guided radiation therapy. In general terms, this means using IMRT with some sort of imaging done immediately prior to treatment to assure that the target is in the expected location. In its simplest implementation, this might involve daily portal films. This approach simply allows us to verify placement of bony structures, and does not truly assure proper placement of the non-bony target. For several years, some departments have used ultrasound systems indexed to the treatment couch, such as the BAT system, for daily localization of the prostate. This is moderately successful, but is not readily applicable to other organs. Much more generally applicable are systems that use either implanted fiducial seeds or daily CT (or both) for guidance. The use of implanted markers is probably the best method when possible. Data from Princess Margaret Cancer Center clearly shows that use of fiducial markers is more reproducible than use of soft tissue contours. This obviously presents some problems since seed placement percutaneously in some locations may be difficult. At present, there are two state of the art types of systems.

 

Tomotherapy™ —

Uses a megavoltage beam on a CT-type gantry both for imaging and for treatment. There are advantages and disadvantages to this approach. The obvious advantage is that the imaging and treatment beams are the same, so there is no problem of registration. However the megavoltage CT's have inherent lack of resolution compared to even the poorest diagnostic scans. In addition, motion artifacts can confound target localization, particularly in the abdomen. The question is whether the images are good enough for the purpose of registration to pie-treatment (planning) CT scans, and the answer is that in most cases they are. These CT images have one great advantage in some cases — they are not sensitive to the atomic weight of the materials traversed, so there is much less artifact from dental fillings, pacemakers, prosthetic joints, etc. However this is very technique dependent. Attempts to get the process done quickly may result in poor results and all the usual difficulties associated with placing IMRT doses in the wrong place. Tomotherapy"' represents a very different way of treating IMRT plans, and holds both great promise and significant pitfalls. There is some concern about increased doses to normal tissues from Tomotherapy, as with all IMRT approaches. This may be important in vulnerable populations (immunosuppressed, pediatric), and importance in other populations remains to be seen. Whether this will result in a higher risk of secondary malignancies in the younger patients is unknown.

 

Linear accelerator with cone-beam CT —

In this approach, used by the Elekta Synergy system, the Varian Trilogy system, and the Siemens Artiste system, a cone-beam CT is attached to the accelerator gantry. The gantry is rotated 360 degrees and a solid state detector mounted directly across from the CT tube is used to collect the images. The resultant images are kilovoltage and therefore have better resolution than megavoltage CT's, but are subject to different but similar artifacts as are seen in megavoltage CVs. Although made with x-ray beams in the kilovoltage range, they do not have the same resolution as diagnostic CT's. As in tomotherapy, patients are scanned in the treatment position and at the time of treatment. IGRT has great promise to reduce set-up errors for IMRT.

 

Radiosurgery —

There are currently three devices that can deliver true radiosurgery: the Gamma Knife, the Cyberknife, and the Novalis system. Radiosurgery has until recently been limited to treatment of lesions that are intracranial or nearly so. Both the Novalis system and the Cyberknife now offer the potential for body radiosurgery. There is no single definition of radiosurgery. Perhaps the following four elements best define radiosurgery as currently practiced:

1. Treatment is directed at the gross lesion as imaged, not including areas at risk (such as nodal areas).
   

2. Treatment is delivered with an accuracy of 2 mm or less.
   

3. Treatment is delivered in a hypofractionated fashion, usually 1-5 treatments.
   

4. Treatment requires either rigid immobilization or imaging during treatment to be sure of targeting. The three available systems use radically different approaches to the provision of radiosurgery.

 

Gamma Knife —

The Gamma Knife uses a fixed array of 201 cobalt sources aligned with great mechanical precision ("0.5 mm) on a single isocenter. Four different collimators are available (4 mm, 8 mm, 14 mm, and 18 mm). Treatment of larger or irregular tumor volumes is achieved by overlapping multiple "shots", ranging in our experience in Oklahoma City from 1-39 (pending recognition by the Guinness Book of World Records). Since the isocenter of the sources is fixed, the patient must be rigidly immobilized, and this is accomplished by application of a titanium frame that is fixed to the skull with four screws that go into the skull. In order to treat different isocenters, the patients head is moved, either by mechanically changing x,y,z settings on bars attached to the frame or by using an automatic positioning system (move the target, not the rifle). A typical treatment cannot be described since there is so much variability. Because frame fixation is required, treatment with more than one fraction is difficult to impossible There are occasional problems with inability to treat very eccentric lesions, particularly in patients with large heads. In patients with smaller heads, it is sometimes possible to treat extracranial sites but there are significant limitations. The accuracy of alignment of the cobalt sources is quite good as mentioned, but there are other sources of set-up error, including frame torsion, MRI distortion, and x,y,z set-up errors. It is estimated that total accuracy is on the range of 1-2 mm. Planning is typically done from MRI images that give high contrast but relatively less accuracy than CT. planning can be done from CT in selected cases. In addition, the system allows for integration of angiographic data, making treatment of AVM's relatively easy and accurate.

 

Novalis —

The Novalis system uses a highly modified linear accelerator with image guidance. It can be used with a frame like that used for the Gamma Knife or can be used in a frameless mode with some loss of accuracy. Unlike the Gamma Knife, it can be used readily for extra- cranial targets. Treatment is delivered by treatment of multiple arcs with the gantry rotating throughout the treatment and the couch assuming different angles between arcs. Imaging is done immediately prior to treatment but not during treatment, leaving the possibility of patient movement during treatment when the system is used in frameless mode. The overall accuracy of the system is on the order of 1-2 mm. The Novalis system suffers somewhat in flexibility due to the use of multiple arcs rather than more flexible beam options. This results in dose distributions that are perhaps less good than those with other radiosurgery systems.

 

Cyberknife —

The Cyberknife uses a small linear accelerator mounted on a highly engineered and accurate robotic arm. This is coupled with orthogonal x-ray imaging that interacts with the robot assembly in such a way that the machine is able to adjust to changes in patient position. Imaging is performed throughout treatment, allowing for constant adjustment to change patient position. Information is sufficient to allow for so-called six dimensional tracking, including translational changes (x,y,z) and rotational changes (roll, pitch, yaw). Accuracy for intracranial treatments is about 0.9 mm. This is preserved for body sites, but degrades to about 1.3 mm when used for moving targets. Because the Cyberknife is used without frame placement, fractionated treatment is possible. This is thought, based on less than adequate data, to be of benefit in situations where dose to critical structures cannot be lowered to a level where single fraction treatment would be deemed safe. The Cyberknife uses a combination of implanted fiducials and LED’S to achieve <1.3 mm accuracy in treatment of moving lesions and is the only system able to do so.