TG-219: Secondary MU Calculations For IMRT and VMAT

Raise your hand if your clinic treats using IMRT or VMAT.

As I suspected, that’s pretty much everyone. It makes good sense why: IMRT and VMAT employ MLCs and sophisticated treatment planning system algorithms to more precisely shape dose than previously possible. And this is precisely why these modalities are as ubiquitous as they are.

While they may be a clinical staple at this point, a task group regarding secondary checks for IMRT/VMAT only recently published a report providing some guidance addressing secondary calculations for these treatment modalities. A decade in the making, TG-219 establishes an approach for IMRT and VMAT secondary dose calculations.

In the span of those ten years required to publish the report, the community has settled into practice patterns of its own. Many of the recommendations may already be common practice as departments have independently developed best practices over time. So what does the new task group bring to the table?

 

TG-219 Basics

If recent chatter on the MedPhys listserv reflects the community at large, there is some skepticism about the value that secondary calculations provide. But as task groups often do, TG-219 methodically lays out the purpose and value of the topic in question. IMRT/VMAT plans are very complex, with a higher potential for errors than conventional 3D planning. Ideally, if errors occur in the planning process (and RO-ILS data shows us they do rather often), interventions to catch them before they reach the patient are key.

These calculations are not intended as a catch-all but instead are designed to add to the integrity of the overall plan verification process. The task group strongly recommends that MU verification should not be used as a replacement for proper patient-specific QA. But having multiple methods in place provides a safety net that apprehends errors when they occur.

But secondary calculations have their limitations. For example, the dose reported to a single point isn’t necessarily representative of the plan as a whole. These calculations don’t catch machine delivery errors. And often, the algorithms used in these checks are more rudimentary than those used in primary calculations.

For that reason, parameters and inputs are often simplified to ensure the process is efficient. This may be changing as commercial options become available, and users adopt more robust and full-scale options. The task group distinguishes between types of algorithms, broadly categorizing solutions into factor-based or model-based approaches:

• Factor-based calculations incorporate values typically used in a hand-calculation, such as PDD/TPR, collimator scatter factor (Sc), etc., to arrive at a dose/MU at the desired point following formalism provided in TG-71.
• Model-based algorithms are more involved, deriving dose to a point or volume based on a more sophisticated approach that involves modeling the multi-source energy fluence from the linear accelerator and converting it to dose using energy-dependent deposition kernels.

The task group also provides recommendations regarding acceptance, commissioning, and implementation of secondary check software. Appropriately, they lean on previous efforts from TG-53, MPPG 5a, and TG-119 to establish best practices on that front. In brief, when bringing on a new software of this kind, they recommend something similar to a conventional TPS, namely a full battery of dosimetric tests, general functionality (import/export, printing, etc.), and treatment geometry and setup. Creating and adequately documenting baselines and repeating tests regularly (and after significant upgrades) helps to ensure the software will continue to serve its function.

 

Recommendations

In summary, the task group converges on the following recommendations, which have been edited for brevity ( see complete recommendations here):

1.  Physicists should not rely only on secondary dose software as a stand-in for IMRT QA.
2.  Secondary MU calculations should be performed for all IMRT/VMAT plans, preferably 2D/3D over 1D.
3.  Algorithms and/or beam data in the secondary check should be independent of the treatment planning system(TPS).
4.  Commissioning of the software should be completed based on the work of TG-53 and MPPG 5a.
5.  Software benchmarking should follow guidelines outlined in TG-119.
6.  Ongoing software checks should be completed annually or after software upgrades.
7.  Agreement between TPS and secondary MU calculations should be within action levels outlined in the report. Acceptability of results should be based on the plan composite and not individual fields.
8.  Evaluate plans that fail to meet acceptability standards so as to understand any discrepancies.
9.  2D/3D calculations for IMRT/VMAT are suggested. Vendors should move away from single-point comparisons.

 

What TG-219 Forgot to Include

Throughout the report, there are various references to existing commercial solutions, providing important and relevant information about the aspects of each software that pertain to secondary calculations. But something is missing from each of these tables and graphs: ClearCalc.

Unfortunate as that may be for ClearCalc users, that shouldn’t come as much of a surprise. After multiple years of work, the finalized task group report was submitted for publication in early 2021. ClearCalc wasn’t released until mid-2020, at which point final report iterations were likely circulating for approval.

Despite its relative youth compared to others in its class, ClearCalc quickly scaled to hundreds of installations in a short time frame. Given the adoption of Radformation’s secondary calculation software, we felt it was important to include additions to the published tables and appendix with ClearCalc-relevant data. This information is intended as a resource for clinics that use ClearCalc that would like to incorporate TG-219 recommendations.

Conclusion

As it stands now, it’s likely the case that we all have a practice in place for validating IMRT/VMAT plans using a secondary calculation platform of some sort. After all, VMAT and IMRT technologies are 15 and 30 years old, respectively. Both quickly became prevalent in radiation oncology departments, and departments had to implement second checks responsibly in the absence of formal guidance. So now that TG-219 published its report, it’s time to evaluate your workflow against new recommendations and fill in the gaps where necessary.