In this second feature of the series on independent QA, Physics World talks to Jeff Kapatoes, senior director for regulatory and research at Sun Nuclear Corporation, a US-based manufacturer of QA solutions for radiotherapy and diagnostic imaging providers, about the necessity for open data access to drive continuous improvements in patient safety, treatment outcomes and workflow efficiency.

A rigorous and independent approach to QA provides an essential audit of the evolving radiotherapy delivery system, complementing the integrated “self-checks” on the treatment machine to ensure that radiation is delivered to the tumour site as intended while minimizing collateral damage to healthy tissues and organs at risk.

How does independent QA benefit the radiation oncology care team?

The clinical medical physicist owns the critical role of independent oversight of patient treatment in the radiation therapy suite. Independent QA providers like Sun Nuclear equip the medical physics team with the specialist QA devices and software they need to fulfil this role effectively – whether that’s commissioning a new treatment machine or clinical workflow; daily, weekly, monthly or annual machine QA checks; as well as all aspects of patient-specific QA, including in vivo QA. For many of these essential checks, open access to the data generated by the treatment delivery system is fundamental to success while ensuring radiation is delivered to the patient as intended.

In terms of specifics, what do you mean by open data access?

Put simply, we’re referring to the availability of treatment delivery data in its broadest sense. For starters, this includes the standard DICOM objects from the treatment planning system (TPS) – the radiotherapy plan, images, structure sets and dose. There are also the cumulative and time-based data from the electronic portal imaging device (EPID) and the machine log files (i.e. monitor units, leaf positions, couch positions, gantry angle and collimator angle). Finally, there are the imaging data – in particular, the cone-beam CT, kV projection and registration offsets.

Is access to all of these data objects guaranteed?

There are notable concerns – in particular, regarding open access to the data generated by the EPID during treatment delivery. While there have already been attempts to restrict access and monetize these data, it’s worth restating that the data are “owned” by the clinic and must remain freely accessible to enable independent analysis of radiotherapy fulfilment.

Right now, though, there are no regulations in the relevant technology standards (e.g. IEC 60601-2-1) mandating the radiotherapy delivery system vendors to provide open data access.

On Elekta’s treatment machines, for example, access to the EPID data is only available to customers via a licence purchase. For Varian, the older C-Series linac allows access, while the newer TrueBeam system only provides access to dosimetric cine data via the service interface. With Varian’s latest Ethos technology, access to the EPID data for patient treatment is not allowed at all, including for Varian’s own portal dosimetry product. This trend is certainly concerning.

"We want data access to be part of the clinical community's collective conversation."

  • Jeff Kapatoes, Ph.D.,
  • Sun Nuclear Senior Director for Regulatory & Research

Why is data access restricted in this way?

Notwithstanding the lack of hard standards requirements, it’s important to acknowledge the non-trivial engineering and manufacturing challenges of architecting a complex treatment system and bringing it to market. Making the associated data objects readily available and transferable for QA purposes takes resource, time and effort – all of which are at a premium when the development team for a new radiotherapy system is focused on shipping to a demanding commercial schedule. As with all functionality, open data access is a matter of prioritization and business decisions.

So open access to the delivery system data is key for an independent and innovative QA ecosystem?

That’s correct. In our opinion, there should be a goal to ensure interoperability is not impeded for competitive reasons but, rather, that it is expected and enabled in a practical manner. The commercial success of independent QA products, in spite of competitive offerings from the established vendors of radiotherapy treatment delivery systems, is the real-world evidence that independent QA is valued by clinical end-users. Bottom line: we cannot take access to these data for granted and must stay vigilant to ensure that access is maintained.

How does open access to data translate into improvements in patient care?

The growing role of automation within the clinical workflow represents a compelling opportunity for efficiency gains. We are now at a point where it’s feasible to carry out in vivo QA of the treatment delivery and better understand what’s happening – monitoring the actual dose delivered to the patient rather than the dose delivered in theory based on the planning images. This, in turn, creates the possibility to do less pretreatment QA, in particular for highly fractionated treatments.

Sun Nuclear’s SunCHECK platform (for integrated machine and patient QA) is part of this evolving value proposition around automation, pulling data objects in automatically, processing those data in the background, organizing the results, and making the care team aware as and when a problem arises. The one-year study of in vivo patient QA conducted by Evy Bossuyt and colleagues at Iridium Kankernetwerk in Belgium is a great example, wherein adjustments to the patient plan were made in 3.7% of the patients treated in that time period. Such changes were enabled via the information provided automatically to the clinical team by SunCHECK.

What about the role of open data access for clinical research?

There’s a lot of activity in research hospitals and universities to figure out how these data objects can be leveraged to deliver enhanced patient safety and workflow efficiency within the clinic. There’s also an acceptance among radiotherapy delivery system vendors – perhaps not publicly stated – of the value to be had in making these data openly available.

The pioneering work of Ben Mijnheer’s team at the Netherlands Cancer Institute (NKI), Amsterdam, is a case in point, yielding significant advances in EPID-based 3D dosimetry on Elekta treatment systems over the past two decades. Meanwhile, Peter Greer and colleagues at Calvary Mater Newcastle Hospital, Australia, have carried out similar research in collaboration with Varian. Their WatchDog system, which uses EPID images to determine treatment delivery accuracy in real-time, has been evaluated across multiple institutions to optimize for error detection.

Longer term, what needs to happen to ensure an open data environment for independent QA checks?

The ideal scenario is a formal stipulation in the relevant technical standards. In other words, a hard-requirement for a system API to give third parties open access to the data objects associated with treatment delivery – information that rightfully belongs to the clinic and the patient. We are willing to participate in any logical manner possible to help make this happen.

For now, our task as an independent QA provider is to encourage clinical users, whenever they’re making a machine purchase from a treatment delivery system vendor, to reiterate the importance of open data access for their medical physics team and their QA programme. In summary: we want data access to be part of the clinical community’s collective conversation with the radiotherapy delivery system vendors.

This article was originally published in Physics World.