Comprehensive end-to-end testing for SRS and SBRT: a practical framework

Stereotactic radiosurgery and stereotactic body radiation therapy have transformed modern oncology by delivering ablative doses with millimetre precision. Centres from Royal North Shore in Sydney to the Peter MacCallum Cancer Centre in Melbourne rely on these techniques for cranial, spinal, and extracranial lesions. Yet that precision is unforgiving: a 1 mm positional error or a 3% dose deviation can mean tumour under-treatment or unintended normal tissue toxicity.

Comprehensive end-to-end testing serves as the bridge between commissioning and routine clinical practice. By replicating the entire workflow, from immobilisation and imaging through planning, image guidance, and beam delivery, physicists detect systematic errors before they reach patients. In Australian departments, where staffing models often require physicists to cover multiple linacs across sprawling networks, this integrated approach offers an efficient safeguard against isolated blind spots.

The framework outlined here reflects lessons learned at metropolitan quaternary centres and regional satellite clinics alike. It draws on the heritage of rigorous quality assurance promoted at meetings such as the COMP 2014 Banff meeting, where international collaboration has long shaped best practice for high-precision radiotherapy.

Defining the scope of end-to-end verification

End-to-end testing differs fundamentally from machine-specific QA. A daily Winston-Lutz check or monthly linac output constancy test verifies individual subsystems, whereas an end-to-end audit exercises every component in sequence. The scope should encompass immobilisation devices, CT simulation parameters, MRI or PET fusion workflows, the treatment planning system algorithm, couch corrections, image guidance registration tolerances, and the final beam delivery itself.

For Australian programmes supporting both public hospital hubs and remote outreach sites such as those servicing the Pilbara or the Top End, defining scope is partly a logistics exercise. A phantom that can be shipped between sites, scanned locally, planned remotely via a central server, and delivered on-site provides a realistic measure of distributed workflows. Establish acceptance criteria before the test begins, and document every step in a checklist that mirrors clinical procedure rather than idealised laboratory conditions.

Phantom selection and imaging protocols

The phantom is the heart of any end-to-end evaluation. For intracranial SRS, anthropomorphic head phantoms with realistic skull and soft-tissue substitutes allow credible CT-to-MRI registration studies. Inserted film, TLDs, or micro-ionisation chambers provide absolute dose measurement alongside spatial resolution. For SBRT, modular phantoms containing lung-equivalent inserts, mobile inserts for motion studies, and high-Z heterogeneities test more complex dose calculations.

Imaging protocols must replicate clinical practice. Use the same slice thickness, reconstruction kernel, and immobilisation mask that a patient would experience. In a Brisbane centre treating liver SBRT, for example, the use of abdominal compression and 4D CT should be reproduced in the phantom study, not bypassed because the phantom is static. This commitment to fidelity distinguishes a true end-to-end test from a dosimetry spot check dressed up as a workflow audit.

Treatment planning considerations

Once the phantom is scanned, the planning exercise should follow the department's clinical protocol as closely as possible. Generate multiple target volumes to test contouring tools, apply heterogeneity corrections, and use the same calculation grid resolution used for patients. Test both static and dynamic conformal arcs, sliding-window IMRT, and volumetric modulated arc therapy if these form part of the institutional SRS or SBRT repertoire.

For end-to-end audits, deliberately introduce known challenges. Place the target near a field edge, near heterogeneity boundaries, or in the high-dose build-up region of the jaws. Evaluate how the treatment planning system reports monitor units and how the record-and-verify system exports them. In Australian clinics running Elekta and Varian platforms side by side, cross-platform testing reveals transfer errors that single-vendor audits can miss.

Small field dosimetry fundamentals

Small field output remains one of the most discussed topics at scientific meetings, where the COMP session referenced as medical-physics-education-in-canada-curriculum-discussions emphasised detector selection for fields below 3 cm. Output factors for these narrow beams require corrections for volume averaging, lateral charged particle disequilibrium, and detector-specific over-response. End-to-end testing offers an ideal opportunity to validate these corrections in a clinical geometry.

Use at least two detector types in the high-dose region: a small-volume ion chamber for absolute dose and a diode or diamond detector for cross-check. Compare against the planning system's calculated dose using gamma analysis with criteria appropriate for stereotactic work, such as 2% and 2 mm with a 10% threshold. Document the results with sufficient granularity that another physicist could reproduce the comparison months later, including ambient temperature and pressure corrections specific to the date of measurement.

Patient-specific quality assurance

Beyond commissioning, every patient plan warrants an individual end-to-end check before the first fraction. This is especially important for multi-isocentre SRS, where misalignment between isocentres can produce clinically significant dose deviations. Measure composite dose delivery on the treatment couch using the actual immobilisation mask, lasers, and image-guidance protocol.

For SBRT sites with respiratory motion, deliver the QA plan with motion enabled if the clinical plan includes motion management. Static QA of a motion-managed plan can mask delivery errors that only appear when the beam follows a moving target. Departments in Melbourne and Adelaide routinely perform these dynamic QA measurements, and the resulting data informs both plan acceptance and ongoing margin policies.

Motion management for SBRT

Motion management adds an entire dimension to end-to-end testing. Whether the strategy is abdominal compression, active breath control, gating, or tumour tracking, each requires validation that the planned dose corresponds to the delivered dose in the presence of realistic motion. Use a programmable motion phantom with known amplitude and period to simulate patient breathing patterns.

Measure the difference between the static and dynamic dose distributions. If the institution treats liver or lung SBRT, the impact of motion on target coverage and organs-at-risk sparing should be quantified for typical respiratory amplitudes seen in the local patient population. ARPANSA guidance and ACPSEM position statements provide Australian benchmarks for acceptable deviations, and end-to-end data should be compared against these national expectations rather than vendor tolerances alone.

Documentation and regulatory compliance

Robust documentation closes the loop. Every end-to-end test should produce a report including phantom specifications, imaging parameters, planning details, delivery parameters, measured doses, calculated doses, deviations, and corrective actions. Align the report structure with the requirements of the Therapeutic Goods Administration for treatment equipment and the National Safety and Quality Health Service standards where applicable.

Retention periods should reflect both regulatory minimums and institutional risk management practice. Many Australian centres retain commissioning records for the lifetime of the equipment, which may exceed 15 years for a high-end linac. Build the documentation template into the quality management system from the outset, so annual end-to-end retests do not require reinvention. The result is a verifiable chain of evidence demonstrating that every patient receiving SRS or SBRT does so on a system whose performance has been independently validated against the clinical workflow.

For Australian medical physicists seeking to benchmark against international peers, the archived materials from international scientific meetings remain a valuable open-access resource. Reviewing historical abstracts and education sessions can sharpen the questions asked during commissioning and refresh approaches to long-standing challenges. End-to-end testing is not a one-off milestone but a continuous discipline, and a strong evidence base keeps that discipline honest. Departments across Perth, Canberra, and Hobart are invited to contribute their end-to-end datasets to inter-centre working groups, because shared evidence lifts the standard of what constitutes acceptable performance nationwide.