The Varian Trans-Energy-Experience™ (TEE)

After TrueBeam was released in 2010 as a groundbreaking new technology, it was clear that there are a few more developments in the pipeline which were unthinkable before, like MLC Tracking, Low MV Imaging, gated Cone-Beam-CT, spiral RapidArc, and others. TEE is one of these developments, and we were able to test it.

(the official Varian TEE™ logo)

Treatment Planning with TEE

The process of treatment planning with TEE is straightforward. The first step is to create a field and select TEE from the drop-down list of energies, where "TEE" is the last item:

From now on, TEE will appear in all places, where an energy step is normally displayed:

As usual, the selected energy is not visible during RapidArc optimization. However, after a plan has been calculated, the Calculation notes reveal the Equivalent Accelerating Potential (EAP), 7.5X in the example:

The preliminary documentation says that the FF-TEE Set for flattened energies uses all three flattening filters of the TrueBeam.

If EAP is between 6X and 8X, the 6MV flattening filter is used. From 8X to 12X, the 10X filter is in the beam. If EAP > 12X, the 15MV filter is active.

This way, beam flattness can be kept within the usual limits, even for continuous TEE energies.

TrueBeam beamlines above S/N 1250 are so-called Continuous Energy (CE) beamlines. The visual identification is easy: besides the "CE" sticker, CE-beamlines are red:

(CE-beamline of TrueBeam STx S/N 1275)

With non-CE beamlines, the red color is a little darker, and more blueish (magenta-like).

What is TEE™?

Everybody knows that the optimal photon energy for a certain treatment plan is never accurately determined. For instance, head & neck patients are always planned with 6 MV photons, whereas for pelvic cases, often the higher photon energy (15 MV, 18 MV) is chosen, based on "tradition" and "experience".

However, it would be quite a coincidence if for a certain h&n planning case the optimal photon energy¹ would just be the preselected 6 MV. It is equally possible, and actually much more likely, that the optimal energy² will be different from 6 MV. Maybe it is 7.1 MV, we would never know. This is where TEE comes into play.

With TEE, the free parameter "photon energy" in the treatment planning process is subjected to optimization. If "TEE" is selected as field energy (see sidebar), the Progressive Resolution Optimizer (PRO) of RapidArc optimization takes care of finding the best energy, which now is a continuous quantity within the waveguide's lower and upper limits, and not one of the fixed energy steps. The resulting optimized energy is called EAP (Equivalent Accelerating Potential). It can only be accessed via the Calculation Notes (see Properties of a treatment field after dose calculation). The field itself will only have the label "TEE" instead of the photon energy (6X, 10X, 15X, ...).

First experiences with TEE

Some other prerequsites exist:

• TEE is only supported by CE-beamlines (see above). If the energy spectrum can't be produced by the linac, it does not make much sense to plan it in Eclipse in the first place.
• The Master file on the TrueBeam has to be edited to unlock the CE features.
• In ARIA, you need a TEE licence (will show up in License Viewer as a site license).
• The server which runs the DCF distributor will get a USB dongle which is branded with the TEE logo. Strange Varian policy.
• Not to forget: Eclipse 11 does not offer TEE support yet. We tried a prerelease of Eclipse 14.

One should also know that TEE currently only works for RapidArc plans, since only the PRO of Eclipse 14 is capable of optimizing energy. It currently neither works for static Gantry IMRT plans nor for 3D-conformal plans. The continuous TEE energy cannot be set manually.

The TEE configuration in Eclipse Beam Configuration is easy. The Master Sets for the main Photon energies (6X, 10X, 15X) have to be measured and configured, as usual. From this FF-TEE Set, Eclipse computes some sort of linear combination of the photon spectra to produce the desired TEE energy.

A separate FFF-TEE Set is generated from the master energies 6FFF and 10FFF. In other words, there is no extra work for the physicist if he chooses to configure TEE in Eclipse. Also the MLC transmissions etc. are interpolated from the Master Transmissions.

From the above it becomes clear that TEE will never extrapolate, but always interpolate. If the lowest photon energy is 6X, the optimizer will never output 5.6X as the optimal energy.

One drawback of TEE exists when dosimetric plan QA is performed (it is more a nuisance than a real drawback). Since some ionization chamber correction factors need the photon energy as input quantity, it is quite cumbersome to look up the EAP in the calculation notes in order to get the corrections right.

Also machine QA is a little more complicated when using TEE. For instance, in order to make a water phantom cross-profile scan at 12.3 MV, one has to use a clinical plan (QA plan) with a high enough MU number, which has been produced by the optimizer of Eclipse. This plan has to be treated on the TrueBeam in QA mode (or Treatment mode), while the water phantom is scanning. Setting a TEE energy directly in Service Mode is not possible (yet).

Conclusions

TEE with the CE beamline is a promising way of getting better plan quality. The key technology is clearly the beamline. Up to now, the linac performed with the same stability for TEE beams as for beams which used one of the master energies.

We wonder what will come next. Maybe Dynamic TEE during a RapidArc treatment, where energy changes during beam-on?

¹We are always talking about an accelerating potential producing a certain photon spectrum. 
²the one which gives the best plan according to the specified constraints.