Per-fraction positional and dosimetric performance of prone breast tangential radiotherapy on Halcyon™ linear accelerator assessed with daily rapid kilo-voltage cone beam computed tomography: a single-institution pilot study

The purpose of this study is to evaluate daily positional and dosimetric accuracy of tangential-field prone WBRT, planned with ECOMP and treated with 6X-FFF beam on Halcyon™ version 2.0 (Varian Medical Systems, Palo Alto, CA). We seek to investigate what parameters affect target coverage and radiation hotspot the most, using Halcyon’s rapid daily CBCT capabilities. This retrospective study was reviewed and approved by our Institutional Review Board.

Image acquisition was done as follows. Eight (8) prone breast patients receiving WBRT post-lumpectomy with hypofractionated schedule of 2.66 Gy × 16 fractions followed by 10 Gy boost were selected sequentially in the 6-month period from a pool of treated patients under IRB approval. There were no exclusion criteria other than ensuring that none of the CBCTs were truncated in a way that would significantly alter dose calculation (e.g. no samples were excluded based on patient characteristics, such as breast separation). CT simulation was performed after patients were immobilized on a commercially available prone breast board. A kV CBCT was taken before delivery of each fraction, as part of Halcyon v2.0 image-guided RT workflow. A total of 16 fx * 8 patients = 128 CBCT datasets were obtained. A synthetic daily CT was generated by deformably registering the planning simulation CT to each of the daily CBCT images using MIM Maestro® (MIM Software Inc., Cleveland, OH) version 6.6. CBCTs associated with boosts were excluded from daily evaluations.

The clinical treatment plans were generated using ECOMP technique (EZfluence™, Radformation, Inc, New York, NY) in Eclipse v15.6 (Varian Medical Systems, Palo Alto, CA), with dosimetric endpoints specified in Table 1. This plan was copied and recalculated on synthetic CT to obtain daily delivered dose distribution. BODY and Breast PTV_eval contours, as defined on Radiation Therapy Oncology Group (RTOG) trial 1005 [14] and census definitions [15] were created on each synthetic CT. The Breast PTV_eval contour was derived for each day of treatment, the process of which is clarified in the following subsections.

Most recent ASTRO guidelines on WBRT state that V105% should be minimized at all times, while some cite 200 cm³ as the recommended limit [12]. The tumor bed is to receive at least 95% of the prescription dose under ASTRO guidance. Similarly, RTOG1005 protocol stipulates (1) 95% of Breast PTV Eval shall receive at least 95% of prescribed dose (D95% > 95%, but D90% > 90% acceptable), and (2) maximum dose less than 115% of prescribed dose (Dmax < 115%, but 120% acceptable).

To investigate if delivered prone WBRT dose on Halcyon meets these guidelines on a per-fraction basis, three primary dose-volume metrics were measured at each fraction to assess PTV Eval coverage and hotspot: V90%, D95%, and V105%. Absolute values of V95% [cc], V100% [cc], and V105% [cc], and global maximum dose (Dmax) were obtained from the BODY contour. Predictive model for these dose-volume metrics were constructed, based on patient parameters including source-to-surface distance (SSD), couch shifts, residual breast position (defined below in Data Collection, Analysis, and Statistics section), and weight.

Eight (8) patients were simulated head-first-prone on Siemens Sensation CT scanner (Siemens Healthineers, Erlangen, Germany) for initial treatment planning. Patients were immobilized with QFix® Prone Breast (Avondale, PA) boards and a vac-lok bag placed underneath the patient and with their arms up. Imaging isocenter was placed midline of the body, at the midpoint (sup-inf) of the breast tissue and anteriorly to the sternum.

WBRT was planned in Varian Eclipse v15.6. Physicians contoured Breast CTV based on consensus definitions [15]. Breast CTV was expanded 5 mm (excluding heart and not crossing the midline) and cropped anteriorly from the skin by 5 mm and posteriorly in front of the rib to obtain Breast PTV Eval. Treatment isocenter was shifted anteriorly to cover Breast PTV Eval. ECOMP technique was carried out by experienced dosimetrists. ECOMP is a forward-planning intensity-modulated radiation therapy (IMRT) technique using parallel opposed beams, where the goal is to deliver as homogenous dose as possible to an irregular surface. For breast, this is done by planning a uniform dose at mid-separation. Skin flash of 2 cm was added beyond the patient contour, and then the edited fluence maps were converted to leaf sequences for Halcyon dynamic MLC. Machine energy was fixed at 6X-FFF (6-MV with flattening filter free), with prescription dose of 266 cGy/fraction for all patients. Anisotropic analytical algorithm (AAA) version 15.6.03 was used for volumetric dose calculation. For photon dose optimization and irregular surface compensator generation, photon optimizer version 15.6.03 was used.

Dose-volume constraints for Breast PTV Eval in a whole breast treatment used at our institution are listed on Table 1. D95% and V90% measure the extent of dose coverage to the breast, while V105% and Dmax are measures of dose hotspots. For this study, a < 200 cm³ objective is additionally applied for the entire patient volume (BODY contour) as another suggested measure of homogeneity by ASTRO consensus but was not a constraint used clinically during planning. Study of normal organs, such as the lungs, hearts, and contralateral breast were omitted from this study.

For treatment on Halcyon v2.0 machines, patients were immobilized with identical QFix® Prone Breast boards and vac-loks used during simulation. Patients were scanned under Halcyon Breast protocol (125 kV, 491 projections at 10 mA and 10 ms, filtered back projection reconstruction) to obtain an on-board CBCT [10]. For 3D/3D matching, chest wall and ribs in the CBCT were aligned with those of the planning image, ensuring the breast in the CBCT is within the planning Breast PTV Eval contour. Online couch shift applied was saved in ARIA for all fractions. Patient body weight was acquired during the weekly on-treatment visit (OTV).

Figure 1a demonstrates the process to generate daily synthetic CTs using MIM Maestro v6.6. In the MIM console, planning CT was deformably registered to the CBCTs to generate synthetic daily CTs. Some synthetic CTs were discarded (2 in total) because of deformation failures stemming from limited CBCT field-of-view. A total of 126 synthetic CT images was extracted from MIM and imported back into Eclipse. Once imported, a copy of original RT plan was re-calculated on each of the synthetic daily CTs keeping all beam parameters intact, while taking into consideration the online CBCT-to-CT registration that was performed during daily treatment. Dose to synthetic daily CTs was calculated with the same AAA version as the original treatment plan.

Figure 1b demonstrates the contouring process on daily synthetic CTs for the treatment target, which is Breast PTV Eval contour. Not all PTV Eval contours were reviewed by physicians, but an effort was made to mimic the original contours from the planning CT in each daily synthetic CTs, by (1) starting with an actual copy of PTV Eval contour from planning CT rigidly registered to the synthetic CT chest wall, (2) modifying the contour to match to chest wall edge shape, and (3) applying a 5 mm subtraction from the breast tissue surface. The breast tissue surface was automatically determined by the Eclipse software in the contouring module for each of the synthetic CTs. Overall, this procedure ensures the PTV Eval structure generated on daily synthetic CT consistently represents the initial physician’s intent on breast tissue being irradiated for each daily fraction.

Dosimetric endpoints of this study include the following. For Breast PTV Eval contour: V90% [%], D95% [%], and V105% [%]. For BODY contour: V95% [cc], V105% [cc], and global Dmax. Here, [%] means “percent of the contour volume” and [cc] means “absolute volume as centimeter cubed (cm³).” Dose-volume histogram (DVH) was calculated using DVH Estimation Algorithm version 15.6.03.

Patient positioning and volumetric data relevant to this study include the following. For general daily positioning data: lateral (LAT), vertical (VRT), and longitudinal (LNG) couch shifts [cm]; source-to-surface distances (SSD) [cm] for each of the two parallel opposed beams; breast width diameter (BWD) derived from SSD, as illustrated on Fig. 1; and post-image-guidance Breast PTV Eval shift [cm] (after accounting for daily couch shifts), referred to as “residual shifts” in this paper, also illustrated on Fig. 1. SSD-derived breast width diameter (BWD) gives a rough estimate of the breast diameter, but also captures some set-up errors of the day resulting in SSD distortions. For example, if both lateral and medial beam SSD increases for the day, BWD decreases; this can be interpreted as either breast diameter decrease or improper breast set-up resulting in breast compression. The residual shift is how much the Breast PTV center-of-mass shifts compared to during CT simulation, after taking into account the couch shifts. For volumetric data: Dice similarity coefficient (DSC) for breast deformations and Breast PTV Eval volume [cc]. DSC measures the extent of overlap between two contours, in our case daily Breast PTV Eval (constructed as shown on Fig. 1b) and planning Breast PTV Eval. DSC is equal to unity when a perfect overlap is achieved, and is calculated as such:Other data collected in this study include machine data: collimator angles, gantry angles, and monitor units for each of the two parallel opposed beams. Patient weight data gathered during simulation and physician visits was spline-interpolated to estimate weight at every fraction.

All data collected were imported into MATLAB (Natick, MA) Version 2019a for data visualization and statistical analyses. Tables and histograms of predictor variables, dosimetric endpoints, and patient position information are presented with descriptive statistics. Non-parametric statistical tests (Wilcoxon signed-rank test) were used to compare daily delivered dose-volumes with dose-volume optimization criteria (Table 1) and with planned dose-volumes. Pearson correlation coefficients among all variables studied are presented, with emphasis on statistical significance of the observed correlation.

Finally, robust parallel slopes model (a type of analysis of covariance or ANCOVA) was constructed [16] to fit Δ dosimetric endpoints using the following predictors: couch shifts, PTV breast residual shifts, Δ BWD, 1-DSC, Δ PTV Eval Volume, and Δ body weight. In this model, each patient is assigned an intercept but the slopes with respect to predictors are assumed to be equal, hence the term “parallel slopes.” The general fitted formula was:Here, \({\text{Y}}\) is the Δ dosimetric endpoints (e.g. ΔV105%), \(\beta_{0i}\) is the intercept for ith patient, \(\beta_{n}\) is the slope for nth predictor \(X_{n}\), and \(\sigma\) is a normally distributed residual term. The individualized intercept \(\beta_{0i}\) for each patient attempts to compensate for patient-specific factors, such as the differences in optimized IMRT fields, machine parameters (gantry angles, monitor units), and other miscellaneous factors. A reasonable estimate of the slopes was made based on this model. Robustness of the model was improved with MATLAB’s default iterative robust least squares procedure using bi-square weights. This reduces effects of high-leverage data and outliers on the fit. Wald’s Test with the null hypothesis that \(\beta_{n} = 0\) estimated statistical significance of each variable on the Δ dosimetric endpoints.

Significance was set at α = 0.05 for most statistical analyses except for multiple tests, for which Benjamini–Hochberg procedure was applied to limit false discovery rate [17].

Table 2 summarizes descriptive statistics of planned and daily patient positioning and contour volume data, as well as difference from prescribed (Δ). Positional data include couch shifts and PTV residual shifts in three orthogonal directions (latitude, vertical, and longitude) and their magnitudes (directional shifts summed in quadrature), source-to-surface distance (SSD) for both parallel opposed (lateral and medial) beams, and DSC comparing planned versus daily target volumes. Daily couch shifts were within ± 2.5 cm in each of the three principal directions, with a mean magnitude of shift of 1.19 ± 0.57 cm. Couch shifts of 0.68 ± 0.40 cm were previously reported in the literature for 21 patients treated on C-arm linac [18]. Breast positioning was consistent, with target center of mass shifting only by 0.26 ± 0.10 cm or 0.30 ± 0.15 cm (PTV Eval and V95% contour center of mass, respectively). The DSC roughly captures breast deformation taking into account the daily couch shifts. At DSC = 0.93 ± 0.02, the intersecting volume of daily and planned contours are 93% of their composite (union) volume and indicates a great set-up consistency. The ΔSSDs of lateral and medial beams were 0.08 ± 0.58 cm and 0.18 ± 0.42 cm from planned, respectively, with some ΔSSD > 1 cm. Patient weights varied from − 6.5% to + 4.7% compared to during CT simulation.

Table 3 tabulates the six delivered dosimetric parameters for homogeneity (Breast PTV Eval V90%, D95%, V105%, BODY V95%, V105%, and global max dose) planned, daily delivered, and difference from prescribed/planned (Δ). Wilcoxon p values indicate the probability that the median value meets the dose-volume objectives (set out on Table 1) for daily measurements or the probability that the median Δ is non-zero. While target coverage was stable, delivered dose had hotspots failing to meet objectives and varying from prescribed dose frequently. The median of daily PTV V105% [%] failed to meet the < 10% objective, at 13.15%. Median global max dose [%] was found to be > 107% at 108.85%, but BODY V105% [cc] was found to be < 200 cm³. Median values of all quantities were within the variation acceptable (see Table 1), but some were not within the treatment objective. Figure 2 shows histograms of daily delivered dosimetric outcomes and difference from prescribed (Δ). The number of treatments that meet the prescribed dosimetric endpoints (i.e. “pass”) or has a small Δ (e.g. “< 10%”) are represented as a fraction out of all treatments (126) examined. The histograms corroborate that target coverage metrics were stable, but dose hotspot metrics were not. Dose objective D95% > 95% was met 111 days out of 126 daily CTs examined, while V90% > 99% was less frequently met at 74/126 (but 95/126 within variation acceptable, V90% > 95%). For difference from prescribed, ΔD95% was > -5% for 123/126 days and ΔV90% was > -1% for 111/126 days. BODY V90% deviated only slightly from planned, with a standard deviation of ± 51.0 cc with a bias of − 16.3 cc. In contrast, PTV V105% < 10% was met only 50/126 days (68/126 within variation acceptable). Dmax was the most often not met, with only 21/126 days meeting the criteria of < 107% of the prescribed dose.

Figure 3 presents PTV V90%, D95%, V105%, BODY V95%, V105%, and global Dmax over the course of 16 fractions as a boxplot. Most metrics deviated from their planned values (indicated as a blue X) with statistical significance indicated within the parenthesis. Measures of coverage (V90%, V95%, and D95%) sometimes decreased during treatment versus planned (Δ < 0). Measures of dose hotspots (V105% and Dmax) generally increased during treatment versus planned (Δ > 0). Some patients, such as patients 2 and 5, experienced both decrease in coverage and increase in dose hotspots.

Figure 4 depicts Pearson correlation matrix among 21 variables studied in this paper, with only statistically significant Pearson coefficients (p > 0.0177) displayed after multiple testing corrections. Negative correlations are colored blue and positive correlations are colored red. The first variable (column or row) represent change with respect to time (fraction number). The next 8 variables represent daily couch shifts, followed by residual shift of the breast center-of-mass. The next 4 variables represent daily patient geometry: dice coefficients (deformation) of Breast PTV Eval contours, SSD-derived BWD, weight, and Breast PTV Eval volume. Remaining 6 parameters are the study endpoints: PTV and BODY dose-volume data. The uppercase Greek delta (Δ) denotes change with respect to planned or simulated (for example, Δ Body Weight [%] means change in patient body weight compared to during simulation).

Correlation of body weight with DSC, BWD, and dosimetric endpoints are plotted in Fig. 5. While some patients lost weight, some maintained weight throughout the course of treatment. Weight loss decreased DSC (or increased breast deformation) and moderately decreased BWD. The change in body weight from simulation (Δ) correlated positively with metrics of dose coverage like PTV ΔV90% (r = + 0.4, slope of 0.071% per weight %) and negatively with metrics of hotspot like ΔV105% (r = − 0.6, slope of − 4.6% per weight %).

The results of comprehensive robust multiple linear regression fit are presented on Additional files 1 and 2. Additional file 1 tabulates the slopes (\(\beta_{n}\)) estimated for each variable. Additional file 2 tabulates the estimated intercepts (\(\beta_{0i}\)) for ith patient and the overall adjusted R² for the model. Each model predicts Δ dose-volume metric. ΔD95% was excluded due to lack of daily plans providing large variations in D95% which resulted in a poor fit (low adjusted R² = 0.4).

Figure 6 summarizes the behavior of residuals and outliers of the robust parallel slopes model described on Additional files 1 and 2. Residuals were mostly normally distributed, but some outliers had residuals deviating more than expected indicating non-linear effects. Outliers for PTV ΔV90% delivered lower-than-expected dose due to simply missing the Breast PTV Eval near the chest wall. Many of these outliers had contralateral breast in the line of irradiation, perhaps indicating there were clinical decisions to avoid irradiating the contralateral breast. Outliers for BODY ΔV105% either had irradiated volume outside of Breast PTV Eval artificially increasing ΔV105% or had extreme variations (> 2 cm) in intra-fraction BWD.