Effect of changes in monitor unit rate and energy on dose rate of total marrow irradiation based on Linac volumetric arc therapy

Total body irradiation (TBI) has long played an important role as a conditioning regimen prior to bone marrow or hematopoietic stem cell transplantation for patients with hematologic malignancies (i.e., leukemia, lymphoma, or multiple myeloma) [1]. The primary reason for the use of TBI is the elimination of residual cancer cells and preventing the immunologic rejection of the transplanted donor stem cells [2]. The TBI treatments are generally performed at an extended source-to-surface distance (SSD) with large open fields to deliver a homogeneous dose to the entire body [3, 4]. In addition, TBI treatments have been performed using techniques such as moving-beam TBI with a sweeping beam, a patient translation technique, multiple-beam TBI, a shrinking-field technique, and the local application of small beams [5‐7].

However, for patients with advanced acute leukemia, increasing the total TBI dose may improve malignant clone killing, but it is also associated with a potential lethal toxicity to the surrounding normal healthy tissues/organs [8, 9]. The side effects of TBI can be extensive and include acute symptoms such as nausea and a loss of appetite, as well as infertility and secondary malignancies [10]. Pulmonary toxicity is one of the most serious, and potentially life-threatening, complications that may occur following TBI [11]. Therefore, new techniques have concentrated on providing a safer dose escalation [12, 13].

To deliver a therapeutic dose to a target, while not irradiating normal tissue, total marrow irradiation (TMI) has been introduced in combination with helical tomotherapy (HT) or volumetric arc therapy (VMAT) using conventional linear accelerators [12, 14, 15]. TMI, when delivered with an intensity-modulation (IM) technique, may reduce the dose delivered to the organs at risk (OARs), relative to TBI. These techniques have the potential to provide a safer dose escalation for a target [16]. Several studies have shown the potential clinical benefits of IM-TMI [12‐15, 17, 18].

For conventional TBI treatments, the dose rate is a significant parameter. To minimize the dose rate, the patient is generally located at an extended SSD of more than 400 cm [3]. Many studies have reported that, when TBI is performed with a low dose rate, the healthy tissue exhibits less damage [8, 9, 11, 19, 20]. Especially, low dose rates (less than 6 cGy/min) can decrease the risk of pulmonary toxicity [11, 19, 21]. However, TMI treatments with HT or VMAT have conventionally been performed with relatively high dose rates due to limitations on the dose rate selection and the treatment location (i.e., at the isocenter). For example, the lowest dose rate is 100 monitor units (MU)/min and 800 MU/min for the Varian C-series Clinac® (Varian Medical Systems, Palo Alto, CA) and HT, respectively. New delivery platforms that can deliver low dose rates have been released by Varian. Their VitalBeam™ (Varian Medical Systems, Palo Alto, CA) and Truebeam™ systems (Varian Medical Systems, Palo Alto, CA) offer dose rates of as low as 5 MU/min and 20 MU/min for 6–10 MV and 15–20 MV, respectively. In VMAT delivery mode, the dose rate can be varied during the treatment. However, the maximum dose rate can be limited by setting the dose rate in the planning process. If, in the plan optimization, the dose rate is set to less than 100 MU/min, which is supported by the C-series, a safe dose rate can be delivered to the OARs, especially the lungs.

In this paper, we describe the plan quality of TMI treatments with a range of MU rates and energies, performed using a Vitalbeam system. The appropriate MU rate setting was determined by evaluating the dose distribution, the dose rate to each OAR, and the duration of the TMI treatment.

The beam-on times and MU for each field, energy, and dose rate are listed in Table 2. Figure 2 shows the non-linear correlation factor between the MUs and beam-on times. The beam-on time decreases as the energy increases. The beam-on time for field 1 (PTV_Brain) was the shortest. The non-linear correlation factor increases dramatically with the MU rate.

Dose-volumetric parameters for each MU rate of TMI by VMAT using 6, 10 and 15 MV was provided in supplementary material (Additional file 1: Table S1).

The TMI isodose distributions in colorwash are illustrated in Fig. 3. The average values of the dose-volumetric parameters of PTV and normal tissue for the energy and dose rate of each patient are provided as supplementary material. The different dose-volumetric parameters for the energy and dose rates are not shown.

Figure 4 shows the dose-rate volumetric histogram for the lung volume of a representative case (Patient 1). When the 40-MU rate was used, the average volume of the lung receiving a dose of less than 6 cGy/min was 92.45 ± 8.88%, 93.47 ± 6.53% and 94.04 ± 6.11% at 6, 10, and 15 MV, respectively. The average lung volumes irradiated at a dose rate of more than 20 cGy/min are listed in Table 3.

For all the plans, the gamma passing rate was more than 90%. For those plans with 300 and 600 MU/min at every energy, the gamma passing rate was 99%. For those plans with 40, 60, and 80 MU/min at every energy, the range of the gamma passing rate was 90–94%.

In the present study, we evaluated the dose rate and dosimetric parameter based on the MU rate of the machine and the energy.

TMI treatments using the VMAT technique are clinically favorable in that they can overcome the limitations of TBI since TMI can reduce the dose administered to an OAR. Additionally, the accuracy of the TMI treatment can be improved using an image guidance system. However, there remain issues with the dose rate when using TMI delivery with a C-series or HT device. However, this problem has been solved by the release of new machines with a low MU rate.

The radiation toxicity of a normal organ, especially the incidence of pulmonary toxicity, is a major factor determining the prescription dose for patients with hematolymphoid malignancies [8, 11, 19, 21]. Pulmonary toxicity can significantly affect the overall survival rate. The dose rate administered to the lung is a key factor, as well as the total dose. The published guidelines recommend that TBI treatments not exceed a dose rate of 20 cGy/min [21]. TBI with a dose rate of less than 6 cGy/min can reduce pulmonary toxicity [11, 19]. When 40 MU/min was set for the PTV_Chest1 and 2 fields, the lung volume receiving less than 6 cGy/min was 90% while the beam-on time of two field (PTV_Chest1 and 2) was 10 min, approximately. However, if MU/min was set for all the fields, the total treatment time would be 60 min. This long treatment time can result in increasing the intra-fraction error (e.g., breathing movement or patient motion. Moreover, the patient can feel discomfort due to lying down on the couch with an immobilization device. To avoid such an adverse effect, we can select the proper MU rate for a specific field which includes an OAR such as the lung. If the maximum MU rate (i.e., 600 MU/min) is applied to the PTV_Brain, Pelvic and Femur fields with 40MU/min for PTV_Chest1 and 2 only, the beam-on time was approximately 24 min. Even if the imaging acquisition time is included, the total treatment can be finished within 1 h. For imaging guide, cone beam computed tomography (CBCT) was taken for each single isocenter. The reference CT image set and digitally reconstructed radiograph was compare to CBCT and pair of orthogonal kV radiographs.

However, in terms of dose voxel, each voxel in OAR has a different dose rate. Each beam of field beam is subdivided into hundreds of beamlets. The voxel was partially irradiated by beamlet which have an individual intensity level [22]. In this study, the dose rate was calculated under the assumption that the dose to OAR is sum total of dose during beam on time.

The total dose administered to the lung did not exhibit any dependency on the energy. However, the beam-on time decreased slightly as the energy increased, except when the dose rate was 600 MU/min. A higher energy can result in the delivery of a similar dose in less time with a limited MU rate. However, when the maximum MU rate (600 MU/min) is allowed, the machine can deliver a sufficiently high dose rate at the maximum gantry speed. The beam-on time with an MU rate of 600 MU/min was constant since the beam-on time is predominantly determined by the maximum gantry speed.

The PTV should include the ribs since the bone marrow is primarily located in the ribs. The lungs, which are the most important OARs, are enclosed by the ribs. In planning the processing, we tried to reduce the dose administered to the lungs as far as possible. However, it was a difficult to reduce this dose due to the PTV coverage including the ribs. The DVH of the lung had a steep dose gradient with a range of 3–10 Gy.

The beam-on time and dose slightly decreased with an increase in the energy. With conventional TBI treatment, high-energy beams (10–18 MV) have historically been used, particularly for patients for which the thickness is greater than 35 cm, and therefore considered to be homogenous. However, when high-energy photon beams of more than 10 MV were used, the photoneutrons can lead to an increase in the undesired dose to the patient [23]. TMI techniques using 10 and 15 MV offer no advantages in consideration of the higher integral dose, larger penumbra and similar beam-on time and dose rate.

The variation in the MU rate was extremely limited relative to the general capacity of the machine. The optimization was processed with the extra constraints for a treatment plan with a lower dose rate. Therefore, plan verification was performed to check the delivery accuracy. The gamma passing rates of the plan with the lower MU rate were lower than those with the higher dose rate. The response of the Si detector used to perform portal dosimetry can be affected by the dose rate. The relatively low gamma passing rate was caused by this effect [24]. However, in every case, the gamma passing rate was more than 90%. All the plans could be delivered clinically.

In this study, the part of the lower extremity (i.e. lower legs from mid-femurs) was not included in VMAT optimization. The current treatment table on the LINAC has a maximum travel length of approximately 150 cm in longitudinal direction [25]. Therefore, two different plans are required for TMI which includes lower leg. In literature, reverse patient position approach has been introduced [18]. However, the field junction problem should be considered carefully for the approach [17]. VMAT plan could be generated for this part in feet-first position, keeping an overlapping region with the upper sector to prevent the risk of cold or hot spots [25].

We evaluated the beam-on time and dose volume parameter for given energies and MU rates for VMAT TMI. The dose volume parameter was found to be not dependent on the energy and MU rate. However, the received PTV dose rate for normal tissue was found to differ significantly with the MU rate. In the case of TMI, the dose rate delivered to the lung is critical. To reduce the probability of normal tissue complications, the selection of the lowest MU rate was recommended for fields included the lung. Considering the total treatment time, maximum dose rate can be selected for other fields.