Partially ablative radiotherapy (PAR) for large mass tumors using simultaneous integrated boost: A dose‐escalation feasibility study

Abstract Purpose This study aimed to assess the feasibility to plan and deliver highly heterogeneous doses to symptomatic large tumors using volumetric modulated arc therapy (VMAT) and simultaneous integrated boost (SIB) during a short course palliative accelerated radiotherapy. Methods A patient with a large symptomatic chordoma infiltrating the right gluteal region was selected. A modified SIB treatment was implemented to irradiate the central volume of the tumor (boost target volume, BTV) up to 10 Gy/fraction in a dose escalation trial while maintaining the remaining tumor volume (planning target volume, PTV) and the surrounding healthy tissues within 5 Gy/fraction in twice daily fractions for two consecutive days. Four SIB plans were generated in the dual‐arc modality; a basal dose of 20 Gy was prescribed to the PTV, while the BTV was boosted up to 40 Gy. For comparison purposes, plans obtained with a sequential boost (SEQ plans) were also generated. All plans were optimized to deliver at least 95% of the prescription dose to the targets. Dose contrast index (DCI), conformity index (CI), integral dose (ID), and the irradiated body volumes at 5, 10, and 20 Gy were evaluated. Results At equal targets coverage, SIB plans provided major improvement in DCI, CI, and ID with respect to SEQ plans. When BTV dose escalated up to 200% of PTV prescription, DCI resulted in 66% for SIB plans and 37% for SEQ plans; the ID increase was only 11% for SIB plans (vs 27% for SEQ plans) and the increase in healthy tissues receiving more than 5, 10, and 20 Gy was less than 2%. Pretreatment dose verification reported a γ‐value passing rate greater than 95% with 3%(global)‐2 mm. Conclusion A modified SIB technique is dosimetrically feasible for large tumors, where doses higher than the tolerance dose of healthy tissues are necessary to increase the therapeutic gain.


| INTRODUCTION
Hypofractionated radiotherapy has been proved to be an effective treatment in local symptomatic control of advanced metastatic tumors, such as pain, bleeding, and obstruction. 1 In the past years, some clinical evidence already suggested the use of shorter fractionation schedules having the same effectiveness as long-course RT in symptom control of patients with incurable cancer. 1,2 For example, a fractionation regimen of twice-daily fractionation (3.70 Gy/fraction twice daily) in 2 days, repeated two times at monthly intervals, was tested in the RTOG 8502 phase II trial for advanced pelvic cancer, 3,4 reporting a significant reduction of grade 3 and 4 late toxicity with respect to traditional treatments based on monthly repeated high single-fraction doses (up to 10 Gy).
As palliative therapy should achieve symptomatic relief with the shortest possible timing, the efficacy of more rapid fractionation schemes was deeply investigated in the last years in our center. In particular, we assessed the tolerability of short-course accelerated RT in twice daily fractions for two consecutive days in several doseescalation trials for head-neck, brain, and pelvic tumors. [5][6][7][8] Our results reported that treatments of twice daily fractions for two consecutive days (5 Gy/fraction) were well tolerated showing a high rate of symptom remission with a good impact on quality of life.
In the last decade, the advances in radiotherapy technology have greatly shown the potential to improve outcomes for patients. In particular, the introduction in clinical practice of intensity-modulated radiotherapy (IMRT) and volumetric modulated arc therapy (VMAT) has greatly improved sparing of normal tissue and, hence, enabling dose escalation and/or intensified fractionation also in the case of large mass tumors. 9,10 Furthermore, IMRT and VMAT techniques allowed the simultaneous delivery of different doses to different target volumes within a single fraction, an approach called simultaneous integrated boost (SIB), that has proven to be more efficient in terms of treatment shortening and radiobiological improved effect. [11][12][13] VMAT demonstrated remarkable capability to explore this flexibility thanks to its rotational delivery modality optimizing incident radiation from 360°a round the patient to achieve highly conformal and heterogeneous doses. 14 In addition, thanks to its rotational dose delivery and to the introduction of flattening filter-free beams (enabling high-dose rate irradiations), VMAT advantages included a large reduction of treatment delivery time, especially for radiosurgical doses. 15 The application of these new advanced techniques to palliative treatments suggests a new possible clinical scenario. Traditionally, standard techniques were aimed to deliver a uniform dose to the tumor volume in order to reduce the tumor size and palliate the symptoms. In cases of large tumors the prescription of a homogeneous dose produces a significant irradiation of the surrounding organs and therefore limits the total dose administrable without serious side effects. But today, the use of advanced techniques as VMAT can allow the use of a dose boost to the central-internal region of the tumor, with the aim to increase tumor response and therefore the palliative effect, without significant increase of healthy tissue irradiation. In particular, if the dose boost is a stereotactic-like dose level, that is, an ablative dose, a partial ablative radiotherapy (PAR) has the potential to amplify both the antineoplastic and palliative effects, not only by improving the cell killing almost in a tumor subvolume, but also by activating immunological antineoplastic mechanisms. 16 This strategy constitutes a modified SIB technique whose main point, as opposed to the widely used SIB technique, is to deliver the highest possible dose to a boost volume, limiting the dose to the healthy tissue to the tolerance dose. 17,18 In this paper we planned a dose-escalation study for a large tumor in the pelvic area to evaluate the feasibility of a short-course accelerated RT based on this modified SIB technique, at five sequential dose levels.

2.A | Patient, volume definition and dose levels
For this dosimetric analysis we selected a patient with a large symptomatic tumor. A 65-year-old female patient with a huge chordoma was enrolled in this study and treated with modified SIB radiotherapy. The patient showed a large swelling infiltrating the right gluteal region and the ipsilateral thigh root (Fig. 1).
The lesion was considered unresectable for its local extension and the presence of lung metastases.  Four plans were optimized with the SIB strategy (SIB20/25, SIB20/30, SIB20/35, and SIB20/40) and generated with the "dualarc" feature, using the optimization process described previously. 9 The planner defines the gantry rotation direction, the start-stop angles and the gantry spacing to define the control points (CP). Then the optimization begins generating coarse initial CP with fluence maps at the start and stop angles and at 24°increments from the start angle, and subsequent MLC sequencing generating two segments per gantry angle. The segments are subsequently then spread out evenly and cloned to achieve the required gantry angle spacing.
All CP are processed to fulfill the motion constraints: maximum leaf speed, dose rates, and delivery times. A direct machine parameter optimization is performed on all CP considering all machine and user constraints, followed by a final dose calculation and segment weight optimization. For all plans, an entire gantry rotation was described in the optimization process by a sequence of 90 control points, that is, every 4°. Collimator was set at 10°to minimize the tongue-andgroove cumulative effect. Dose calculation was performed using the collapsed cone convolution algorithm and a dose-grid resolution of 2 × 2 mm 2 in the axial plane.
For comparison purposes, plans obtained with a sequential boost (SEQ20/25, SEQ20/30, SEQ20/35, and SEQ20/40) were generated using one arc for each target. All plans were optimized to deliver at least 95% of the prescription dose to the PTV. For the BTV, the goal was to deliver the boost dose to at least 95% of the volume, penalizing volumes receiving more than 107% of prescription dose.
In particular, a priority goal was to enhance the steepness of dose gradient outside the target volume. In MasterPlan optimization module, this task can be effectively performed using the so-called "surrounding dose fall-off" objective. 19 This objective is able to control the rate of dose fall-off within a structure (e.g., the patient body), penalizing doses above a certain level at a certain distance to the target. In other words, this function is a linearly decreasing dose level starting at the high dose level adjacent to the target and dropping to the low-dose level at a defined distance from the target. In this study, the surrounding dose fall-off objective was used in order to potentially decrease the dose from 20 to 10 Gy in a 4-cm distance. was equal to 30 Gy; the BED to the PTV25, PTV30, PTV35, and PTV40 were equal to 40.6, 52.5, 65.6, and 80 Gy, respectively. The BED contrast was defined similarly as the ratio of BED to the BTV and to the PTV. The normalized BED contrast (nBDC) was defined similarly as the ratio of actual BED contrast and ideal BED contrast.

2.C.2 | Conformity index
Conformity indexes (CIs) were defined as the volume encompassed by the 95% isodose divided by the PTV volumes. 21 A CI value closer to 1 indicates a more conformal dose distribution to PTV.

2.C.3 | Integral dose and healthy tissue irradiation
Integral dose (ID) is the volume integral of the dose deposited in the patient body and is equal to the mean dose times the volume irradiated to any dose (excluding the PTV). 22 The ID was then used to evaluate the cost to deliver highly heterogeneous dose to the BTV.
This was reported together with the irradiated volumes at the dose levels of 5, 10, and 20 Gy (V5, V10, and V20).

| RESULTS
All plans satisfied the target dose coverage objectives. Figure 2 shows the dose contrast indexes of SIB and SEQ plans as a function    To better quantify the dose distribution differences between SEQ and SIB plans, Table 1 reports the absolute volumes of the isodose clouds receiving 5 Gy, 10 Gy, 15 Gy, and 19 Gy (the 95% of prescribed dose at basal level).
An accurate pretreatment verification was performed for three plans of increasing complexity (the basal plan, SIB20/30, and SIB20/ 40). Each arc was delivered two times, once in coronal and then in sagittal plane, so that each plan has four measurements. Table 2 shows the mean and range of the gamma passing rate (γ%) for each plan obtained using different acceptance criteria (2%/2 mm, 3%/ 2 mm 3%/3 mm) with both local and global normalization. When the commonly used 3%/3 mm criterion was considered, γ% was above 98% and 90% for all plans using global and local normalization, respectively. The γ% was observed to decrease as the criteria became more stringent. In addition, there is a clear tendency that the gamma passing rate decreases with increasing plan complexity, for example, as the boost dose increases. However, also with the stringent 2%(global)/2 mm criterion, the mean passing rate achieved an acceptable value of at least 94% for all plans. Modern radiation therapy techniques have also allowed non conventional approaches to obtain highly heterogenous doses within a large tumor volume. This technique, called "Lattice", represents an evolution of the older high-dose GRID radiotherapy 31 and allows for localized 3D high-dose array within the tumor volume, in which highly concentrated hot-spot doses are located in lattice vertices with a rapid dose fall-off between them, then resulting in a periodic three-dimensional peak-to-valley dose distribution. 32 Recent research in radiobiology has provided a theoretical basis using the concept of bystander effect within the parts of irradiated tumor volume not directly irradiated. 33 In addition, a robust abscopal effect in distant tumors or metastatic lesions that are not directly treated have been also described. 34 All these data strongly suggest that the use of very high heterogeneous doses could induce a higher rate of tumor cell apoptosis in bulky and hypoxic tumors than conventional radiotherapy. An excellent agreement (more than 95%) between the measured and calculated dose distributions for the 3%(global)/2 mm criterion was found for all plans, regardless of plan complexity. These results comply the suggestions of the recent AAPM task group no. 218 report, 41 focused on tolerance limits and methodologies for IMRT measurement-based verification QA. This report recommended as universal tolerance limits a γ-passing rate ≥95%, with 3%/2 mm criterion, a 10% dose threshold and using global normalization in absolute dose. However, the utility of common adopted γ-passing rate for the purpose of patient-specific dose QA has been recently questioned, since this method may lack sensitivity and specificity in predicting clinically important patient dose errors. 42 We then performed a more stringent γ-analysis, restricting traditional tolerances; despite stricter discrimination in terms of dose difference/distance to agreement, our results confirm the deliverability of our modified SIB-VMAT technique and its reliability and safety for clinical applications.

| CONCLUSIONS
We introduced a modified SIB technique having the potential to be particularly effective for large bulky tumors, where doses higher than the tolerance dose of healthy tissues are necessary to increase the therapeutic gain. We showed that despite the major dose escalation in the boost volume, the dose conformity to PTV and the integral dose to the normal tissue minimally increased, with a dose spillage from PTV to normal tissue almost constant. The pretreatment dose verification supplied an excellent agreement with calculated values ensuring the accuracy of delivered dose distribution in clinical cases.
The safe delivery of ablative dose in the central part of the tumor is feasible and has the potential to greatly improve the palliative effect.

CONFLI CT OF INTEREST
The authors declare no conflict of interest.