Development and validation of an automatic commissioning tool for the Monte Carlo dose engine in myQA iON

Independent dose verification with Monte Carlo (MC) simulations is an important feature of proton therapy quality assurance (QA). However, clinical integration of such tools often generates an additional and complex workload for medical physicists. The preparation of the necessary clinical inputs, such as the machine beam model, should therefore be automated. In this work, a methodology for automatic MC commissioning has been devised, validated, and developed into a MATLAB tool for the users of myQA iON, the recent QA platform of IBA Dosimetry. With this workflow, all necessary parameters can easily be tuned using dedicated optimization methods. For the geometrical beam parameters (phase space), the assumption of a single or double Gaussian is made. To model the energy spectrum, a Gaussian function is assumed and parameters are optimized using either MC simulations or a library of pre-computed Bragg peaks. For the absolute dose calibration, commissioning fields can be reproduced with the dose engine to retrieve the necessary parameters. We discuss in a first time the tool efficiency and show that one can optimize all parameters in less than 4 min per energy with excellent accuracy. We then validate a beam model obtained with the tool by simulating homogeneous spread-out Bragg peaks (SOBPs) and patient QA plans previously measured in water. An average range agreement of 0.29 ± 0.34 mm is achieved for the SOBPs while 3%/3 mm local gamma passing rates reach 99.3% on average over all 62 measured patient QA planes, which is well within clinical tolerances.

Infarct characteristics and outcome of patients with transient ST-segment elevation myocardial infarction compared to ST-segment and non-ST-segment elevation myocardial infarction

Background

Up to one out of four patients with signs of ST-segment elevation myocardial infarction (STEMI) express complete normalization of ST elevation before primary revascularization procedure. This condition is commonly referred to as “transient ST-segment elevation myocardial infarction” (TSTEMI) and recent data suggests that this group of patients may have favorable outcome compared to STEMI patients. However, it is currently unknown how these patients compare to both STEMI and non-ST-segment elevation myocardial infarction (NSTEMI) patients with respect to infarct size characteristics and outcome.

Objective

This study aims to explore cardiac magnetic resonance (CMR) derived scar tissue and 1-year outcome in patients with TSTEMI by comparison to STEMI and NSTEMI.

Methods

Patients with STEMI were enrolled from two prospective studies (n=170); the patients with TSTEMI were recruited from the TRANSIENT trial (n=141); the patients with NSTEMI were prospectively and consecutively collected from local registries of Amsterdam UMC (n=57) and Maastricht UMC (n=51). All patients underwent CMR examination 2–8 days after the index event. Cine imaging was done for volume and function assessment. Late gadolinium enhancement imaging was performed to identify infarct size (in grams) and the presence of microvascular obstruction (MVO). All CMR images were processed in a single core laboratory (Amsterdam UMC). Clinical outcome after 1 year was measured by the incidence of major adverse cardiac events (MACE), defined as recurrent myocardial infarction (MI), revascularization and all cause death.

Results

The TSTEMI group demonstrated the lowest end-systolic left ventricular volume and highest left ventricular ejection fraction across the groups (overall p<0.001). Although there was a remarkably lower infarct size in TSTEMI patients compared to STEMI (1.41g [0.00–3.91] vs 13.48g [5.31–26.81], p<0.001), there was only a trend towards lower infarct size compared to NSTEMI patients (1.41g [0.00–3.91] vs 2.13g [0.00–8.64], p=0.06). Whilst MVO was observed less frequently in TSTEMI compared to STEMI patients (5 (4%) vs 53 (31%), p<0.001), no significant difference was seen between TSTEMI and NSTEMI patients (5 (4%) vs 5 (5%), p=0.72). Multivariable linear regression analysis identified infarct type, smoking, peak troponin-T and pre-PCI TIMI flow as predictors for infarct size (p=0.03, p=0.03, p<0.001 and p<0.001, respectively). One-year mortality rate was low in all 3 MI types (TSTEMI 3 (2.2%), NSTEMI 3 (3.1%), 4 (2.4%), log-rank test p=0.91). However, there was a significant difference in MACE at 1 year across the 3 MI types (TSTEMI 18 (13.2%), NSTEMI 19 (19.4%), STEMI 11 (6.7%), overall p<0.01).

Conclusion

In comparison to NSTEMI and STEMI, TSTEMI yielded favorable cardiac left ventricular function and scar mass. However, this did not lead to benefit in short term (1-year) outcome; further studies are needed with longer follow-up.

Funding Acknowledgement

Type of funding source: Other. Main funding source(s): European Association of Cardiovascular Imaging (EACVI) Research Grant

Impact of machine log-files uncertainties on the quality assurance of proton pencil beam scanning treatment delivery

Irradiation log-files store useful information about the plan delivery, and together with independent Monte Carlo dose engine calculations can be used to reduce the time needed for patient-specific quality assurance (PSQA). Nonetheless, machine log-files carry an uncertainty associated to the measurement of the spot position and intensity that can influence the correct evaluation of the quality of the treatment delivery. This work addresses the problem of the inclusion of these uncertainties for the final verification of the treatment delivery. Dedicated measurements performed in an IBA Proteus Plus gantry with a pencil beam scanning (PBS) dedicated nozzle have been carried out to build a 'room-dependent' model of the spot position uncertainties. The model has been obtained through interpolation of the look-up tables describing the systematic and random uncertainties, and it has been tested for a clinical case of a brain cancer patient irradiated in a dry-run. The delivered dose has been compared with the planned dose with the inclusion of the errors obtained applying the model. Our results suggest that the accuracy of the treatment delivery is higher than the spot position uncertainties obtained from the log-file records. The comparison in terms of DVHs shows that the log-reconstructed dose is compatible with the planned dose within the 95% confidence interval obtained applying our model. The initial mean dose difference between the calculated dose to the patient based on the plan and recorded data is around 1%. The difference is essentially due to the log-file uncertainties and it can be removed with a correct treatment of these errors. In conclusion our new PSQA protocol allows for a fast verification of the dose delivered after every treatment fraction through the use of machine log-files and an independent Monte Carlo dose engine. Moreover, the inclusion of log-file uncertainties in the dose calculation allows for a correct evaluation of the quality of the treatment plan delivery.

Towards clinical application: prompt gamma imaging of passively scattered proton fields with a knife-edge slit camera

Prompt γ-ray imaging with a knife-edge shaped slit camera provides the possibility of verifying proton beam range in tumor therapy. Dedicated experiments regarding the characterization of the camera system have been performed previously. Now, we aim at implementing the prototype into clinical application of monitoring patient treatments. Focused on this goal of translation into clinical operation, we systematically addressed remaining challenges and questions. We developed a robust energy calibration routine and corresponding quality assurance protocols. Furthermore, with dedicated experiments, we determined the positioning precision of the system to 1.1 mm (2σ). For the first time, we demonstrated the application of the slit camera, which was intentionally developed for pencil beam scanning, to double scattered proton beams. Systematic experiments with increasing complexity were performed. It was possible to visualize proton range shifts of 2-5 mm with the camera system in phantom experiments in passive scattered fields. Moreover, prompt γ-ray profiles for single iso-energy layers were acquired by synchronizing time resolved measurements to the rotation of the range modulator wheel of the treatment system. Thus, a mapping of the acquired profiles to different anatomical regions along the beam path is feasible and additional information on the source of potential range shifts can be obtained. With the work presented here, we show that an application of the slit camera in clinical treatments is possible and of potential benefit.

Feasibility and robustness of dose painting by numbers in proton therapy with contour-driven plan optimization

PURPOSE

To prove the ability of protons to reproduce a dose gradient that matches a dose painting by numbers (DPBN) prescription in the presence of setup and range errors, by using contours and structure-based optimization in a commercial treatment planning system.

METHODS

For two patients with head and neck cancer, voxel-by-voxel prescription to the target volume (GTVPET) was calculated from (18)FDG-PET images and approximated with several discrete prescription subcontours. Treatments were planned with proton pencil beam scanning. In order to determine the optimal plan parameters to approach the DPBN prescription, the effects of the scanning pattern, number of fields, number of subcontours, and use of range shifter were separately tested on each patient. Different constant scanning grids (i.e., spot spacing = Δx = Δy = 3.5, 4, and 5 mm) and uniform energy layer separation [4 and 5 mm WED (water equivalent distance)] were analyzed versus a dynamic and automatic selection of the spots grid. The number of subcontours was increased from 3 to 11 while the number of beams was set to 3, 5, or 7. Conventional PTV-based and robust clinical target volumes (CTV)-based optimization strategies were considered and their robustness against range and setup errors assessed. Because of the nonuniform prescription, ensuring robustness for coverage of GTVPET inevitably leads to overdosing, which was compared for both optimization schemes.

RESULTS

The optimal number of subcontours ranged from 5 to 7 for both patients. All considered scanning grids achieved accurate dose painting (1% average difference between the prescribed and planned doses). PTV-based plans led to nonrobust target coverage while robust-optimized plans improved it considerably (differences between worst-case CTV dose and the clinical constraint was up to 3 Gy for PTV-based plans and did not exceed 1 Gy for robust CTV-based plans). Also, only 15% of the points in the GTVPET (worst case) were above 5% of DPBN prescription for robust-optimized plans, while they were more than 50% for PTV plans. Low dose to organs at risk (OARs) could be achieved for both PTV and robust-optimized plans.

CONCLUSIONS

DPBN in proton therapy is feasible with the use of a sufficient number subcontours, automatically generated scanning patterns, and no more than three beams are needed. Robust optimization ensured the required target coverage and minimal overdosing, while PTV-approach led to nonrobust plans with excessive overdose. Low dose to OARs can be achieved even in the presence of a high-dose escalation as in DPBN.

Unusual presentation of a mesenchymal eyelid hamartoma and an update of the incidence of periocular hamartomas in dogs

A mesenchymal hamartoma in the dorsomedial eyelid of a Staffordshire bull terrier and the incidence and histological features of twenty-two canine mesenchymal periocular hamartoma cases are reported. The archives of the “Comparative Ocular Pathology Laboratory” of Wisconsin (USA) were searched for canine mesenchymal periocular hamartoma. Signalment, clinical appearance, location and histological findings are summarized for twenty-two dogs, containing fourteen different breeds, between four and fourteen years old. Fifteen hamartomas were located at the lateral canthus. Histologically, they consisted of fully differentiated fibrous tissue interspersed with adipose tissue, with bundles of skeletal/smooth muscle in ten cases, and peripheral nerve tissue in two cases. No mitotic figures were noted. Mesenchymal hamartomas may present as a subcutaneous, subconjunctival or orbital mass. Although they have a predisposition to occur at the lateral canthus, they may be located elsewhere on the eyelids or in the orbit.