Introduction
The treatment of patients with multiple brain metastases usually involves whole brain radiotherapy (WBRT) or stereotactic radiosurgery (SRS). WBRT has been replaced by SRS and stereotactic radiotherapy (SRT) since WBRT tends to cause neurocognitive deterioration. Reference Chea, Fezzani and Jacob1
SRS is a single-fraction high-dose treatment technique for brain metastases. Typically, SRS is used to treat small, well-defined intracranial tumours that are not adjacent to critical organs. In case of either larger tumours or tumours that are close to critical organs, SRS poses a higher risk of causing brain radionecrosis or damage to critical organs near the target; therefore, a smaller dose-per-fraction, multi-fraction treatment, SRT, is often the preferred method of treatment. SRS and SRT treatments are both non-invasive or minimally invasive compared to surgery and enable high doses to be administered to tumour volumes, with sharp dose fall-offs at the borders of tumour(s) with submillimetre accuracy and high conformity. Reference Ma2 SRS/SRT treatments could be delivered with linac-based (photon), gamma-knife, cyber-knife and proton therapy.
Traditionally, SRS for multiple brain metastases has been delivered with multiple isocentres. For linac-based delivery, using conventional volumetric arc therapy (VMAT), this is typically achieved with one isocentre per lesion. Reference Agazaryan, Tenn and Lee3–Reference Blomain, Kim and Garg5
HyperArc is a new SRS treatment delivery technique (Varian Medical Systems) that allows the treatment of multiple brain lesions by assigning a single isocentre to treat a set of scattered lesions. Reference Moleme, Hau Chun Wong, Ali and Mugabe6,Reference De Ornelas, Diwanji and Monterroso7 Unlike conventional multi-isocentric VMAT treatments, HyperArc enables a mono-isocentric non-coplanar VMAT plan and simplifies the delivery of SRS treatments by automating the process, whereby eliminating the need for the therapists to manually execute couch shifts to the next field position during the delivery of each field. The overall process is more efficient with minimal treatment time and an improved dose distribution. Reference Van Herk, Remeijer, Rasch and Lebesque8–Reference Juda, Fillmann and Kukolowicz11 The HyperArc treatment planning process is also more automated compared to a traditional conventional VMAT planning. HyperArc treatment planning employs a class solution template that determines the optimal common isocentre location, collimator angles and non-coplanar beam arrangements and minimizes the workload for the dosimetrists. Reference Ohira, Ueda and Akino12
A mono-isocentre HyperArc plan will function as effectively as a conventional brain SRS plan in the case of clustered lesions. However, significant dosimetric discrepancies may exist when the distance between individual lesions increases or as the size of the lesions gets smaller. In other words, when the lesions are far away from each other, small shifts in the common isocentre may have more effect on the dose coverage, especially for the smaller tumours. Reference Sagawa, Ohira and Ueda13
The planning target volume (PTV) in radiotherapy is defined as the clinical target volume (CTV) plus a margin to account for uncertainties and errors during treatment. These errors are mainly setup errors (systematic and random errors) and patient intra-fractional motion errors, which occur due to patient movement or organ motion during the treatment. Reference Zhang, Shan and Liu14 It is imperative to select the correct PTV margin for HyperArc treatments since small lesions are treated with the same isocentre, and choosing an inappropriate margin could result in missed targets and potential overdose of organs at risk. Reference Ohira, Komiyama and Kanayama15,Reference Fung, Wong and Lee16
Nevertheless, there is no consensus on the optimal PTV margin for HyperArc SRS plans. It typically varies between zero and 1 mm. Reference Meeks, Mercado and Popple17 Some centres use zero PTV margin by setting PTV equal to gross tumor volume (GTV); Reference Popple, Brown and Thomas10 some add 1 mm margin to CTV to obtain PTV, Reference Moleme, Hau Chun Wong, Ali and Mugabe6,Reference Ohira, Ueda and Akino12 and others use 1–2 mm margin. Reference Agazaryan, Tenn and Lee3 The most widely used margin prescription formula is Van Herk’s formula of 2·5Σ + 0·7σ, where Σ stands for the standard deviation of group systematic error and σ is the standard deviation of random error. Reference Ohira, Ueda and Akino12 The systematic errors include setup and patient positioning errors, whereas random errors stand for patient intra-fractional motion errors and inter-fractional motion errors. However, Herk’s formula does not provide any insight into how the size of the tumours and the separation between them should affect the choice of margin. The choice of PTV margin for HyperArc SRS plans ultimately depends on the size of the tumours and the separation between them Reference Shen, Wang and Shao18 and for varying combinations of these parameters. There is a need for a quantitative guideline of intra-fractional shift versus target coverage data for varying combinations of tumour size and separation.
The effect of beam modelling errors on dose delivery has been studied extensively. Reference May, Hardcastle and Hernandez19 Our study focuses specifically on the effect of intra-fractional motion on target volume and dose coverage and analyses the limits of zero-PTV-margin treatments. The effect of setup error on dosimetric performance is analysed for larger tumours (diameter of 1 cm and greater); Reference Hisashi, Satoshi and Satoru20 we are interested in the effect of errors emerging due to intra-fractional shifts on percent dose and volume coverages of targets with diameters ranging from 3 mm to 1 cm in this study.
Our purpose in this paper is to examine the effect of lesion size and lesion separation on dose coverage for HyperArc SRS plans, as well as to provide a comprehensive, quantitative guideline that will assist in determining the optimal margin needed to account for intra-fractional motion. This study outlines the limits of safe zero margin treatment and states under what circumstances (lesion size, separation and setup error) zero-PTV-margin treatment still accomplishes the desired target dose and target volume coverage.
Materials and Methods
Retrospective study of registered shifts
This study aims to provide quantitative data on how patient intra-fractional motion influences dose coverage for HyperArc stereotactic SRS treatments. Our starting point was to determine the historical range of intra-fractional shifts at our centre by performing a retrospective study. An analysis of post-cone beam CT (CBCT) shifts of previous HyperArc brain SRS patients was conducted for this purpose. The retrospective study included 42 patients and 159 fractions. Figures 1 and 2 summarize retrospective study data on historical shifts in our centre. The translational intra-fractional shift in the isocentre is broken down into vertical, longitudinal and lateral shifts. The rotational intra-fractional shift in the isocentre is broken down into Pitch, Roll and Rtn (Yaw).
Retrospective data for the percentage of translational shifts for patients who had HyperArc brain SRS treatments obtained via post-CBCT scans. The data includes 42 patients and 159 fractions. The black horizontal line stands for the average of vertical, longitudinal and lateral shifts for each shift interval.
Retrospective data for the percentage of rotational shifts for patients who had HyperArc brain SRS treatments obtained via post-CBCT scans. The data includes 42 patients and 159 fractions. The black horizontal line stands for the average of rotational shifts for each shift interval.
The mean and standard deviations for the vertical, longitudinal and lateral shifts for the population (159 fractions) were 0·3 ± 0·3 mm, 0·3 ± 0·2 mm and 0·2 ± 0·3 mm, respectively. The mean and standard deviations for the rotational shifts Pitch, Roll and Yaw for the population were 0·31 ± 0·31°, 0·17 ± 0·21° and 0·24 ± 0·29°, respectively. The mean and standard deviation of the population show that there is a significant spread in the data.
As can be seen from Figure 1, at our institution, on average, 93·3% of the translational shifts were less than 0·8 mm, 89·7% were less than 0·6 mm and only 2·1% were greater than 1 mm. Similarly, Figure 2 shows that, on average, 94·7% of all rotational shifts were less than 0·8°, and 90·7% were less than 0·6°. Only 1% of the shifts were beyond 1·0°. Note that these results include outliers. From our retrospective data, the relevant range of shifts at our centre is 0–1·4 mm translational and 0–1·2° rotational.
After determining the range of translational and rotational shifts to be tested, we need to employ a practical method to measure the effect of shifts on dosimetric performance. Testing the effect of translational shifts on target coverage is more practical than rotational shifts since one can shift the original plan along x, y or z directions and calculate the dose with the original plan monitor units in the Eclipse Treatment Planning System (TPS) but cannot do the same with rotations.
As seen in Figure 3, a rotation about the isocentre for a certain separation of two tumours corresponds to a specific translational shift for each tumour. Tumour A in Figure 3 translates to a new location as a result of a rotational shift of ɵ degrees.
Tumour shift corresponding to a rotation of angle ɵ.
Each relevant translational shift corresponding to a specific rotational shift can be calculated from Figure 3. Hence, we generated the translational shifts corresponding to relevant rotational shifts as listed in Table 1.
The table lists the calculated shifts (in centimetres) corresponding to angle of rotations ranging from 0·2 to 1·2 °. Shifts are rounded to two decimal points. Tumour to isocentre distance is half the distance between targets
Rotational shifts were converted to translational shifts for different tumour separations (Table 1). Resulting shifts were applied to the original plan, and the dose was calculated with the original plan monitor units. This provides us with quantitative information on percent dose and volume coverage for the target to allow for the estimation of the effect of rotational and corresponding translational shifts on target coverage.
Simulation study
The STEEV Phantom was CT scanned with an Encompass mask with a Siemens Somatom Sensation Open CT scanner and imported into the Eclipse TPS, version 16·01·10. Two solid volumetric identical spherical contours (GTV1 and GTV2) are generated on Eclipse TPS version 16·01·10. Volumetric contours are set to high resolution. The energy and dose rate for SRS treatment were set to 10 FFF and 2400 Mus/min, respectively. HyperArc plans were constructed such that both GTV1 and GTV2 received 1800 cGy. Calculation model and optimization were set to AcurosXB_1610 and PO_1610, respectively. Structure resolution and calculation resolution are set to high resolution (1·25 mm) and 1 mm, respectively. For optimization of the HyperArc plan, jaw tracking and ALDO (automatic lower dose objective) were set active. Targets were selected as GTV1 and GTV2, and target weights were set to 100.
For each reference plan, the target projection margin optimization parameter was set to small (1–2 mm) in the Eclipse TPS. Particularly as the target size and the amount of shift applied decrease, this becomes increasingly important. Setting the target projection margin setting to small results in a tight fit between the 99·5% isodose line and the target for the reference plan with no rotational shift. Any shift applied to the target will thereby result in a reduction in dose and volume coverage. As a result, the effect of the intra-fractional shift applied to the plan can be determined more sensitively.
The initial unshifted plans were normalized to V100% = 99·5% for each tumour separation and tumour size. We then applied a translational shift to the original plan and calculated the dose using the same monitor units as the original plan without normalizing it. As an example, for an 11 cm separation, the translational shift that corresponds to 0·2° rotational shift is 0·02 cm, as shown in Table 1. We repeated the calculations for the same lesions with varying distances between them to determine the effect of rotational shifts on the dosimetry as the separation between lesions increases. The separations between lesions ranged from 5 cm to 15 cm as shown in Table 1. Additionally, we changed the sizes of the lesions and recalculated the dose coverage to better understand how the rotational shifts affected the dose coverage based on the size of the lesions. Lesion diameters used in this study were 0·3 cm, 0·4 cm, 0·6 cm, 0·8 cm and 1·0 cm. Using interpolation, we could estimate the approximate values for other separations.
It should be noted that applying the translational shift in any chosen direction may result in changes to the target coverage of individual tumours, but the combined target coverage (GTV Total) would be ideally unaffected. However, creating tumours with a diameter less than 5 mm does not always result in perfectly symmetrical volumes. In this regard, the choice of shifting direction might result in an increase in the standard deviation of data for smaller targets. To minimize the effects of subtle imperfections in tumour shape and their effect on dose and volume coverage, we carried out simulations with equal shifts in the -x, -y and -z directions separately and considered the average coverage in all three directions as the ultimate value for the percentage dose and volume coverage.
The tabulated data tables (Tables 2–6) list the percent dose and volume coverage for both individual targets (GTV1 and GTV2) as well as the combined target (GTV Total) for varying rotational shifts and separations. The listed data is the mean percent dose and volume coverages along -x, -y and -z directions. The standard deviations of each data are also listed in the tables.
a) The minimum percent dose received by 100% of the tumour volume. GTV1, GTV2 and GTV Total represent the first target, second target and combined target volumes, respectively. Each D100% value in the table has been averaged out separately for identical shifts along -x, -y and -z directions. Table 1 summarizes translational shifts that correspond to rotational shifts. b) Listed in the second table are the standard deviations of the average percent dose coverage values (D100%)
a) The minimum percent dose received by 100% of the tumour volume. GTV1, GTV2 and GTV Total represent the first target, second target and combined target volumes, respectively. Each D100% value in the table has been averaged out separately for identical shifts along -x, -y and -z directions. Table 1 summarizes translational shifts that correspond to rotational shifts. b) Listed in the second table are the standard deviations of the average percent dose coverage values (D100%)
The absolute volume of healthy brain that receives 1200 cGy and more. Volume is in cubic centimetres
The absolute volume of healthy brain that receives 1200 cGy and more. Volume is in cubic centimetres
The absolute volume of healthy brain that receives 1200 cGy and more. Volume is in cubic centimetres
a) The minimum percent dose received by 100% of the tumour volume. GTV1, GTV2 and GTV Total represent the first target, second target and combined target volumes, respectively. Each D100% value in the table has been averaged out separately for identical shifts along -x, -y and -z directions. Table 1 summarizes translational shifts that correspond to rotational shifts. b) Listed in the second table are the standard deviations of the average percent dose coverage values (D100%)
a) The minimum percent dose received by 100% of the tumour volume. GTV1, GTV2 and GTV Total represent the first target, second target and combined target volumes, respectively. Each D100% value in the table has been averaged out separately for identical shifts along -x, -y and -z directions. Table 1 summarizes translational shifts that correspond to rotational shifts. b) Listed in the second table are the standard deviations of the average percent dose coverage values (D100%)
a) The minimum percent dose received by 100% of the tumour volume. GTV1, GTV2 and GTV Total represent the first target, second target and combined target volumes, respectively. Each D100% value in the table has been averaged out separately for identical shifts along -x, -y and -z directions. Table 1 summarizes translational shifts that correspond to rotational shifts. b) Listed in the second table are the standard deviations of the average percent dose coverage values (D100%)
Data Analysis and Results
We analysed the effect of shifts up to 1·2°, which is the most extreme case according to our patient records, on target coverage. We plotted the target volume and target dose coverage (V100% and D100%) of 3 mm versus 10 mm diameter tumours for an extreme shift of 1·2°.
Figures 4 and 5 illustrate the effect of 1·2° shift on percent volume and dose coverage for 3 mm and 10 mm tumours, respectively, for varying lesion separations from 5 cm to 15 cm.
Percent dose (D100%) and percent volume (V100%) coverage of 3 mm tumours for increasing lesion separations.
Percent dose (D100%) and percent volume (V100%) coverage of 10 mm tumours for increasing lesion separations.
As shown in Figure 4, the effect of shift is significant for small tumours and large separations. For a maximum separation of 15 cm, the percent volume coverage, V100%, was 39·1%, while the percent dose coverage, D100%, remained significantly higher at 79·6% for 3 mm tumours.
In the case of larger tumours, the effect of shift on the percent dose and volume coverage is less significant. Figure 5 shows that for a maximum separation of 15 cm, the percent volume coverage, V100%, is 80·2%, while for a tumour of 10 mm, the percent dose coverage, D100%, is 77·1%.
The tabulated data attached to this paper (Tables 2–6) summarize the percent volume and dose variations for all separations and shifts. The data also contains the absolute volume of the healthy brain that receives 1200 cGy for the original plans. A total tumour separation of 11 cm is more clinically relevant than 15 cm. 11 cm separation corresponds to a 5·5 cm distance from the isocentre of two identical GTVs. For two 3 mm diameter tumours with 11 cm separation (5·5 cm from the isocentre) for an extreme shift of 1·2°, D100%, V100% and V12Gy[cc] were 84·2%, 56·1% and 0·45cc, respectively (Table 2a–e). Similarly, for two 4 mm diameter tumours with 11 cm separation (5·5 cm from the isocentre), for an extreme shift of 1·2 °, D100%, V100% and V12Gy[cc] were 80·3%, 66·8% and 0·57cc, respectively (Table 3a–e). For two 6 mm diameter tumours with 11 cm separation (5·5 cm from the isocentre) for an extreme shift of 1·2 °, D100%, V100% and V12Gy[cc] were 78·6%, 75·0% and 0·71cc, respectively (Table 4a–e). In addition, for two 8 mm diameter tumours with 11 cm separation (5·5 cm from the isocentre) for an extreme shift of 1·2 °, D100%, V100% and V12Gy[cc] were 78·6%, 81·1% and 1·34 cc, respectively (Table 5a–e). For two 10 mm diameter tumours with 11 cm separation (5·5 cm from the isocentre) for an extreme shift of 1·2 °, D100%, V100% and V12Gy[cc] were 82·1%, 85·6% and 2·32cc, respectively (Table 6a–e).
c) Percent volume that receives 100% of the total dose. GTV1, GTV2 and GTV Total represent the first target, second target and combined target volumes, respectively. Each V100% value in the table has been averaged out separately for identical shifts along -x, -y and -z directions. Table 1 summarizes translational shifts that correspond to rotational shifts. d) Listed in the second table are the standard deviations of the average percent volume coverage values (V100%)
c) Percent volume that receives 100% of the total dose. GTV1, GTV2 and GTV Total represent the first target, second target and combined target volumes, respectively. Each V100% value in the table has been averaged out separately for identical shifts along -x, -y and -z directions. Table 1 summarizes translational shifts that correspond to rotational shifts. d) Listed in the second table are the standard deviations of the average percent volume coverage values (V100%)
c) Percent volume that receives 100% of the total dose. GTV1, GTV2 and GTV Total represent the first target, second target and combined target volumes, respectively. Each V100% value in the table has been averaged out separately for identical shifts along -x, -y and -z directions. Table 1 summarizes translational shifts that correspond to rotational shifts. d) Listed in the second table are the standard deviations of the average percent volume coverage values (V100%)
c) Percent volume that receives 100% of the total dose. GTV1, GTV2 and GTV Total represent the first target, second target and combined target volumes, respectively. Each V100% value in the table has been averaged out separately for identical shifts along -x, -y and -z directions. Table 1 summarizes translational shifts that correspond to rotational shifts. d) Listed in the second table are the standard deviations of the average percent volume coverage values (V100%)
c) Percent volume that receives 100% of the total dose. GTV1, GTV2 and GTV Total represent the first target, second target, and combined target volumes, respectively. Each V100% value in the table has been averaged out separately for identical shifts along -x, -y and -z directions. Table 1 summarizes translational shifts that correspond to rotational shifts. d) Listed in the second table are the standard deviations of the average percent volume coverage values (V100%)
Discussions and Conclusions
According to our simulations, dosimetric performance (D100%, V100%) is strongly dependent upon tumour separation and tumour diameter parameters. Percent dose coverage (D100%) and percent target volume coverage (V100%) are both affected by varying distances between tumours and varying lesion sizes. However, even in the most extreme cases, D100% remains greater than 74·1%, while V100% drops to 39% (Tables 2–6).
At our centre, 97·4% of all shifts were less than 1·2°, 94·7% less than 0·8° (Figures 1 and 2). When the target-to-isocentre distance is within 3·5 cm (7 cm separation), D100% is over 90% coverage. Over 90% coverage for zero PTV margin, even for an extreme shift of 1·2°, is achievable as long as the target-to-isocentre distance is within 3·5 cm. Similarly, 90·7% of shifts are within 0·6°. For 11 cm separation (5·5 cm from the isocentre) for 3 mm tumours, the D100% and V100% are 93·8 and 88%, respectively. A percent dose coverage over 93% is achievable with zero margin for a 0·6° shift (Table 2a). For larger tumours, both percent volume and percent dose coverage improved significantly (Tables 3–6). For all scenarios, the healthy brain dose constraint of V12Gy<8cc was satisfied.
Using this study as a quantitative guideline, clinicians can decide which PTV margins to use for HyperArc brain SRS treatments. In this study, we applied the translational shifts (Table 1) to eclipse plans that corresponded to specific rotational shifts and listed the results as rotational shifts versus D100% and V100% (Tables 2–6). The effect of rotational and translational shifts on dose coverage for a particular lesion size and separation can be estimated by referring to Table 1 and Tables 2–6. Depending on the specific scenario, a more stringent PTV margin may be adopted if the healthy brain dose is high. In addition, the dosimetric consequences of adding or not adding a PTV margin can be deduced, leading to a more accurate estimation of dose coverage and healthy brain dose.
As an example, for an 11 cm separation, a 1·0° rotational shift corresponds to a 1 mm translational shift according to Table 1. For 3 mm tumours, a 1 mm translational shift brings D100% down to 88% from 99·3%. As a result, we are able to determine how the dose coverage would be affected if we were to proceed with zero margin and were to encounter a 1 mm intra-fractional motion during treatment. When tumours have a larger diameter and/or a closer proximity, the penalties associated with the percent dose and volume coverage are less severe. Due to the linear trend in the data, it is also possible to interpolate the uncalculated distances and diameters. To illustrate, for 3 mm tumours, if a 1 mm shift causes an 11·3% decline in D100%, bringing it down to 88% from 99·3%, then a 0·5 mm shift or error will cause, on average, a 5·65% difference, resulting in D100% falling to 93·7%.
This study outlines the limitations of zero-PTV-margin treatments that ensure target volume and dose coverages (V100% and D100%) are above 90% while the total brain satisfies V12Gy<8cc. Although single isocentre multi-target radiosurgery has been studied extensively, Reference Chea, Fezzani and Jacob1,Reference Meeks, Mercado and Popple17,Reference Shen, Wang and Shao18,Reference Hisashi, Satoshi and Satoru20 our study highlights the limitations of zero-PTV-margin treatments for as small as 3 mm diameter tumours and provides a quantitative guideline for various scenarios in the form of tabulated data, to guide the oncologist to estimate the dosimetric performance for a specific diameter of tumour with varying distances from the common isocentre.
In our retrospective study that is specific to our centre, we isolated and analysed the effect of uncertainty due to intra-fractional motion on dosimetry. However, the source of uncertainty does not play an important role in estimating the percentage dose coverage reduction. Any uncertainty could be applied to the original plan as a rotational shift, and the same monitor units could be delivered to the shifted plan to analyse the impact of uncertainty on dose coverage and volume coverage. In addition, our data could be interpolated to estimate the dosimetric effect on targets of custom sizes and separations. A study by Hisashi Nakano et al. Reference Hisashi, Satoshi and Satoru20 evaluated the effect of setup error on the reduction of dose coverage for targets with a diameter ranging from 1·0 cm to 3·0 cm. As part of our study, we examined smaller targets with diameters ranging from 3 mm to 1·0 cm. In Hisashi’s analysis, 0·5° and 1·0° of rotational errors would result in 2·3 and 4·6% reduction of dose coverage, respectively, for a 1·0 cm diameter target 3·0 cm from the isocentre. Based on our data (Table 6a), we estimated that dose reductions for the same 0·5° and 1·0° of rotational shifts (errors or uncertainties) for 1·0 cm diameter targets are 2·4 and 6·1%, respectively, by interpolating our data.
Tables 2–6 summarize average percent dose (D100%) and percent volume (V100%) coverages for the first target (GTV1), second target (GTV2) and the combined target volume (GTV Total). All values in the tables are averaged out over -x, -y and -z directions. The applied translational shifts that correspond to rotational shifts are listed in Table 1. The standard deviations for all values are listed in separate tables. For 3 mm targets, the standard deviations are higher compared to larger targets. This is because the number of shifts applied is comparable to the size of the targets. Also, small imperfections in tumour shapes for smaller tumours yield higher discrepancies in dose coverages in different directions. If the targets (GTV1 and GTV2) were perfectly symmetrical, one would expect the same dose reduction for all shifts and separations. GTV Total values have a narrower spread with smaller standard deviations. For this reason, it is preferable to use the GTV Total values as a reference to estimate the dose reduction for an arbitrary distance from the isocentre for an arbitrary target size.
Our data shows that the V12Gy dose is strongly correlated with the target size. For a maximum rotational shift of 1·2°, the maximum V12Gy healthy brain dose was 0·51 ± 0·01cc (Table 2e) for 3 mm diameter targets, whereas it was 2·40cc ± 0·02cc (Table 6e) for 10 mm targets. Besides, target volume and dose coverage decrease with increasing distance from the common isocentre. Similar trends have been observed in various publications. Reference Ohira, Komiyama and Kanayama15,Reference Fung, Wong and Lee16,Reference Golmakani, McGrath and Williams21
The absolute volume of healthy brain that receives 1200 cGy and more. Volume is in cubic centimetres
The absolute volume of healthy brain that receives 1200 cGy and more. Volume is in cubic centimetres
In this study, uncertainties may arise from subtle geometrical and volumetric imperfections associated with generating targets of smaller diameters. Creating contours of 3 mm might not always result in identical shapes and volumes. Our method for minimizing this effect was to apply the same shift in three directions (x, y, z) and average the results. An additional potential uncertainty is the sensitivity of shifts in the Eclipse TPS. In Eclipse TPS, shifts in centimetres are allowed to be applied to three significant figures. Therefore, the geometrical shifts calculated in Table 1 are rounded to three significant figures in centimetres. For targets of smaller sizes, such as 3 mm and 4 mm, these uncertainties might result in a 0–3% difference in percent dose and volume coverage.
In this study, nearly perfect spherical targets were used, but it should be noted that in real clinical scenarios, targets are not perfectly symmetrical. This study however demonstrated that the larger the diameter of the targets, the better the percent dose and volume coverage. Results from our study could be utilized to estimate dose coverages in various situations.
It is crucial that HyperArc is utilized effectively in the clinic. The conventional VMAT multi-target plans may increase the total brain dose in the event of metastasis. The HyperArc treatment reduces this risk by treating multiple tumours around the common isocentre, thereby reducing the number of fractions for brain SRS treatments. This guideline aims to optimize the quality of HyperArc SRS plans in the clinic.
Open access
