Background
Stereotactic Body Radiation Therapy (SBRT) emerged from stereotactic radiosurgery (SRS), where successful treatments of brain tumours were delivered using a single high dose fraction with a steep dose falloff [
1‐
4]. Therefore, the earlier definition of SBRT focussed on using a stereotactic body frame for patient setup, which allowed conformal dose distribution to be delivered to an extracranial target in a hypo-fractionated treatment scheme [
5]. Now, image guidance techniques are preferred and can provide a frameless method for patient setup. Current definitions by various organisations indicate that SBRT is a form of external beam radiotherapy, in which a high dose of radiation is accurately delivered to an extracranial target in a small number of fractions [
6‐
9].
SBRT is commonly used to treat primary cancers of the lung, prostate, kidney, liver, and pancreas [
10] and oligometastatic cancers in the lung, liver, lymph nodes, adrenal gland, and spine [
11]. This review focuses on the treatment of early-stage non-small cell lung cancer (NSCLC) using SBRT. Global data from 2020 shows that lung cancer has been the second most commonly diagnosed cancer and responsible for the highest number of cancer deaths [
12]. NSCLC comprises 85% of lung cancer diagnoses with about 15% at a localised early stage [
13,
14], for which conventionally, surgical resection has been the standard treatment [
15]. However, patients may be deemed medically inoperable, due to a high possibility of treatment complications associated primarily with poor pulmonary function, or other factors such as performance status, cardiovascular disease, age and comorbidities. SBRT has replaced conventional radiotherapy as the standard treatment for inoperable early-stage NSCLC patients [
16]. Local control (LC) rates for 3- and 5- year follow up periods range between 78 − 98% and 79 − 85%, respectively for early-stage NSCLC when treated with SBRT [
14]. This is a significant improvement compared to LC rates for treatment with conventional radiotherapy, which for 2- and 5- year follow up periods, were typically around 40% and 10%, respectively [
16].
Considerable research on factors affecting patient outcomes for SBRT of the lung have been accumulated over the past few decades, including image-guided radiotherapy (IGRT) and motion management [
17], patient conditions [
18], fractionation schedules [
15] and tumour size [
15]. Due to the complex nature of SBRT, the required immobilization, verification, and treatment delivery lead to considerably longer treatment times [
19,
20]. In addition to patient discomfort, longer treatment times have the potential to increase intrafraction patient motion, which would be unsuitable for the required tight margins for SBRT treatments [
21]. The speed of Volumetric Modulated Arc Therapy (VMAT) delivery is generally limited by the maximum dose rate imposed by the flattening filter. While removal of the flattening filter increases dose rate and decreases treatment time, it results in a non-uniform dose profile [
20]. Intensity modulation is used to compensate for this; however, it raises the question: Is the dose distribution of modulated FFF beams equivalent to that of flattened beams?
The key is to achieve a good therapeutic ratio along with minimal intrafractional motion, which is especially important in the given case as the treatment involves very high doses delivered in a few fractions. This review aims to investigate dosimetry for photons beams with or without flattening filter, as well as various photon energies, to establish the optimal choice for lung SBRT, particularly in terms of factors such as target coverage, sparing of critical organs, treatment delivery time and the number of monitor units (MUs) used.
Discussion
In SBRT treatments, it is essential to maintain an accurate treatment position as higher doses are delivered in fewer fractions leaving less leeway for error. The BOT needs to be reduced to minimise the intrafractional variation, along with ensuring a good therapeutic ratio.
Lung SBRT is associated with high tumour control rates for both single-fraction [
41] and multifraction treatments [
41‐
43]. In the case of single-fraction SBRT, flattened 6 MV and 6-FFF VMAT resulted in equivalent dosimetric parameters for PTV and OARs; and BOT was significantly reduced with the use of 6-FFF beams as expected [
23]. However, Pokhrel et al. [
28,
39]. suggested that 6-FFF improved target conformity, dose coverage at tumour interface and OAR sparing. This is more prominent for lower ipsilateral lung density [
28]. Lu et al. [
36]. recommended using 6-FFF for 3 × 18 Gy and 4 × 12 Gy schemes, and 10-FFF for 1 × 34 Gy scheme. The tumour reportedly remains in a considerably stable position if the treatment delivery time is 6 min or less [
44,
45]. This implies that tumour motion is not the primary concern in the case of 3 × 18 Gy and 4 × 12 Gy schemes (BOT in the range 1.5-4 min), and they would benefit the most from 6-FFF beams, which provides better OAR sparing. The BOT for 1 × 34 Gy scheme was 6.3 min with 6-FFF, and 3.5 min with 10-FFF, indicating that 10-FFF would be beneficial for this fractionation scheme.
The flattening-filter free technique can be used without compromising the target coverage [
22‐
26]. Due to the lower beam energy of FFF beams, there is less scattering in linear accelerator head, and less MLC leakage and scattering in the medium. As a result, there is a sharp dose gradient outside the PTV. This can be beneficial in improving the beam conformity as well as reducing the OAR dose [
25,
27,
28]. FFF beams can considerably lower the NTCP for critical organs [
26,
36]. Although FFF plans are advantageous for OAR sparing, it is difficult to achieve such good results in the case of larger tumours [
26]. Vieillevigne et al. [
29]. compared DCA or VMAT plans using various flattened and FFF beams, and showed that particularly for DCA, the conformity and healthy-tissue sparing was suboptimal when 10-FFF was used for medium and large targets. Meanwhile, the results from Wu et al. [
35]. suggested that the OAR sparing improved with FFF beams for large targets. The studies reviewed in the present article showed that differences in target coverage, CI, HI, and OAR sparing between the flattened and FFF beams are generally small and may not be clinically significant.
There are relatively fewer studies that investigate dosimetry for different energies of photon beams, for example, 6-MV, 10-MV, and intermediate megavoltage beams (< 6-MV). Based on the literature comparing 6-FFF,10-FFF and flattened 6 MV beams, it is indicated that all provide comparable dose distributions to the target [
31,
32,
34]; however, 6-FFF can significantly improve conformity and OAR sparing [
22,
29,
31,
32,
34]. This is attributed to sharper penumbra at shallow depths and small fields, caused by the shorter secondary particle ranges, which is particularly advantageous for NSCLC treatments [
32]. 10-FFF is found to be more beneficial for skin sparing as the maximum dose shifts further away from the surface in comparison to 6 MV beams [
34]. It must be noted that in the case of lung SBRT, skin dose is only a concern when the tumour lies very close to the skin. This is a clinically rare situation, and no high-grade skin toxicity has been observed in patients treated with 6-FFF [
32]. Zhang et al. [
46]. suggested that 3-MV can be a potentially better choice for treatment of patients who are physically thin, as it could further improve the tumour coverage and OAR sparing in comparison to 6-MV beams.
Using FFF beams can reduce the BOT by a factor of 2.3 compared to flattened beams, also compensating for the increase in MUs with FFF [
22,
23,
28,
29,
32,
34]. With reduction in treatment time, a subsequent reduction of intrafractional motion is observed, which enables the use of smaller PTV margins [
38]. The BOT reduction is slightly greater when 10-FFF is used instead of 6-FFF [
32,
34,
36]. Depending on the case examined, the improvement in treatment efficiency can outweigh the small increase in OAR dose [
34]. Another factor to consider is the interplay between the target motion and the motion of the photon beam defined by the multi-leaf collimator (MLC) aperture which can affect the accuracy of dose delivery. When the BOT is shorter, the interplay effect becomes a concern and robust optimization may be required to counter it [
23]. However, it is practically negligible for treatments using 2-arcs and more than 2 fractions [
20].
Of note, type-B dose algorithms, such as AAA (Varian Medical Systems, Inc., RRID:SCR_017372) and Collapsed Cone (Philips, RRID:SCR_008656) have been implemented in various treatment planning studies included in the review instead of type-C dose algorithms, such as Acuros XB (Varian Medical Systems, Inc., RRID:SCR_017372). While comparing AAA with Acuros XB, AAA overestimates the dose quite significantly in the low-density regions of the lung (≤ 0.15 g/cm3) and this effect increases for small tumors (≤ 15-mm diameter) and in the case of 10-MV photon beam [
47]. Dose calculations performed with Acuros XB are more accurate than AAA or Collapsed Cone, as well as in good agreement with the X-ray voxel Monte Carlo calculations [
48].
It is important to incorporate patient-specific factors in the decision making when selecting the optimal beam modality for the treatment. If the tumour is not at a sufficiently large distance from the skin, 10-FFF would be implemented to avoid skin reactions [
34]. The feasibility of the photon beam should be verified with respect to the tumour size [
29]. The weight of the patient would also have an impact. As discussed earlier, the use of intermediate MV photon beams might be useful for physically thin patients [
46]. In patients with cardiac implantable electronic device (CIEDs), FFF beams with energies greater than 6 MV would not be ideal if the tumour is close to the device [
49]. A lower energy, 6 MV, is better suited for patients of older age, with a poorer ECOG performance or a higher Charlson comorbidity index.
Results from previous reviews by Dang et al. [
20]. and Ghemis et al. [
50] are congruent to the results obtained in the present review, regarding the benefit of FFF-SBRT in reducing the beam-on time along with sufficient tumour control and OAR sparing. The current work extends this analysis to also include comparison of FFF beams with 6 MV or 10 MV energies, providing clinically applicable evaluation of the associated dosimetric effects. The robustness of the results was limited by the lack of literature than compares 6 MV and 10 MV for lung SBRT. Other limitations that need to be addressed are smaller patient group, inadequate use of dose verification tools and few planning studies performed with type-C dose algorithms. Potential investigation of patient outcome with assessment of local control, overall survival, acute and late toxicity, etc. can also be carried out.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.