Dosimetric evaluation of incorporating the revised V4.0 calibration protocol for breast intraoperative radiotherapy with the INTRABEAM system

Abstract In breast‐targeted intraoperative radiotherapy (TARGIT) clinical trials (TARGIT‐B, TARGIT‐E, TARGIT‐US), a single fraction of radiation is delivered to the tumor bed during surgery with 1.5‐ to 5.0‐cm diameter spherical applicators and an INTRABEAM x‐ray source (XRS). This factory‐calibrated XRS is characterized by two depth‐dose curves (DDCs) named "TARGIT" and "V4.0.” Presently, the TARGIT DDC is used to treat patients enrolled in clinical trials; however, the V4.0 DDC is shown to better represent the delivered dose. Therefore, we reevaluate the delivered prescriptions under the TARGIT protocols using the V4.0 DDC. A 20‐Gy dose was prescribed to the surface of the spherical applicator, and the TARGIT DDC was used to calculate the treatment time. For a constant treatment time, the V4.0 DDC was used to recalculate the dosimetry to evaluate differences in dose rate, dose, and equivalent dose in 2‐Gy fractions (EQD2) for an α/β = 3.5 Gy (endpoint of locoregional relapse). At the surface of the tumor bed (i.e., spherical applicator surface), the calculations using the V4.0 DDC predicted increased values for dose rate (43–16%), dose (28.6–23.2 Gy), and EQD2 (95–31%) for the 1.5‐ to 5.0‐cm diameter spherical applicator sizes, respectively. In general, dosimetric differences are greatest for the 1.5‐cm diameter spherical applicator. The results from this study can be interpreted as a reevaluation of dosimetry or the dangers of underdosage, which can occur if the V4.0 DDC is inadvertently used for TARGIT clinical trial patients. Because the INTRABEAM system is used in TARGIT clinical trials, accurate knowledge about absorbed dose is essential for making meaningful comparisons between radiation treatment modalities, and reproducible treatment delivery is imperative. The results of this study shed light on these concerns.


| INTRODUCTION
Breast intraoperative radiotherapy (BIORT) is the delivery of radiation to the tumor bed to treat neoplastic cells within the surgical margin at the time of surgery. 1 The delivery of intraoperative radiotherapy (IORT) is made possible by mobile linear accelerators that produce electrons (3)(4)(5)(6)(7)(8)(9)(10)(11)(12), high-dose Ir-192 after loaders, or low-energy kilovoltage (kV) x-ray generators such as the INTRA-BEAM ® (Carl Zeiss Surgical GmbH, Oberkochen, Germany) or Xoft Axxent ® (Xoft Inc. is a subsidiary of iCad Inc., San Jose, CA.). This study focuses on the dosimetry of the INTRABEAM x-ray generator, which is primarily used to deliver adjuvant BIORT for early-stage breast cancer and is shown to be a viable alternative to whole breast irradiation (WBI) through the results of the targeted intraoperative radiation therapy A (TARGIT-A) randomized trial. 2,3 The advantage of BIORT is reduced treatment length; enhanced patient convenience; and reduced dose to the contralateral breast, heart, and lungs. [3][4][5] The efficacy of BIORT is being further investigated with ongoing national clinical trials. The TARGIT-US registry trial (ClinicalTrials.gov Identifier: NCT01570998) is studying the efficacy and toxicity of breast INTRABEAM IORT with or without WBI.
Clinical trials such as these rely on the accurate knowledge of dose distribution to perform meaningful dosimetric comparisons between radiation treatment modalities.
The INTRABEAM system is factory-calibrated annually and characterized by two-calibrated depth-dose curves (DDCs) named "TAR-GIT" and "V4.0." The TARGIT DDC gained its name from its use within the TARGIT-A clinical trial. 6 The TARGIT DDC is acquired  7 The PTW 23342 and PTW 34013 chambers differ with exposure and air-kerma calibration schemes, respectively.
Because of their cavity volume differences, these two chambers require distinct chamber holders and have different effective points of measurement. Consequently, at depths proximal to the x-ray source (XRS), the TARGIT and V4.0 DDCs report differences in dose rate. 8 Because greater differences in dose rate are shown at depths proximal to the XRS and a smaller spherical applicator surface is more proximal to the XRS, there is a greater concern of dose accuracy with the use of smaller spherical applicator sizes. The TARGIT DDC is maintained to ensure prescription consistency between TAR-GIT-A and current clinical trials (i.e., TARGIT-B, TARGIT-E, TARGIT-US), while the V4.0 DDC is used to perform calibration consistency checks with the Zeiss water phantom.
When compared with film and ion chamber measurements, the TARGIT DDC underestimates the delivered dose by 14-80%, while the V4.0 DDC difference ranges 1-5%. 7 Even though the TARGIT DDC underestimates the delivered dose, it is used in treatment delivery for patients enrolled in TARGIT clinical trials. The purpose of this study is to reevaluate the dose, dose rate, and equivalent dose in 2-Gy fractions (EQD2) for patients enrolled in TARGIT clinical trials using the V4.0 DDC.

2.A | INTRABEAM system
The INTRABEAM system consists of a miniature XRS attached to a counterbalanced-mobile-floor stand with six degrees of freedom.
This stand helps position the XRS within the patient for radiation treatment delivery. When a 50-kV accelerating voltage is applied across the x-ray tube, a beam of electrons is accelerated through a drift tube (10.0-cm length and 0.32-cm diameter) toward a thin gold target to produce x-rays. The vendor of the INTRABEAM system reports its x-ray beam to have these qualities: 50-kV, 20.4-keV effective energy (E eff ), and 0.64 mm of aluminum (Al) half-value layer (HVL) at 1-cm depth in water. 9 To treat the lumpectomy cavity, a rigid-water-equivalent plastic spherical applicator, within the range of 1.5-to 5.0-cm diameter, is attached and centered to the x-ray probe and then inserted into the lumpectomy cavity. 10 For the ≤ 3.0-cm diameter spherical applicators, an Al filter is incorporated into the spherical applicator design to remove low-energy photons from the treatment spectrum. 11 This Al filter increased treatment delivery time because it reduced the dose rate at the surface of the spherical applicator. 10 The bodies of the > 3-cm diameter spherical applicators remove low-energy photons from the treatment spectrum; thus, they do not require a filter. 12

2.B | Calibration consistency check
We performed a calibration consistency check of the manufacturerprovided calibrated V4.0 DDC for our XRS and used a self-shielded Zeiss water phantom. Dose rate measurements with this water phantom were directly compared with the V4.0 DDC. This phantom features a precise three-dimensional translational stage, which offers reproducible source mounting and allows the XRS to be translated, with a precision of 0.01 cm, relative to the chamber. 7 Within the phantom, the PTW model 34013 ionization chamber can be supported by two fixed waterproof chamber covers: one is for isotropy, and the other is for depth-dose measurements. A DDC is generated when the XRS is translated away from the ionization chamber and a charge reading is collected with an electrometer.
The PTW model 34013 ionization chamber is calibrated with traceability to Physikalisch-Technische Bundesanstalt (PTB) in Germany using a reference x-ray beam with an E eff = 16.4-keV and HVL = 0.43-mm Al, which is often referred to as "T30." As suggested by the vendor and endorsed by other investigators, this beam quality is best matched to the INTRABEAM spectrum. 9,13 The American Association of Physicist in Medicine (AAPM) endorses PTB as an approved primary standards dosimetry laboratory and approves the use of this calibration approach for the INTRABEAM system. 14 For depth (z), the manufacturer suggests that the measurement of dose SHAIKH ET AL. | 51 rate in water for the XRS DR W-XRS (z) can be expressed in Equation The ionization charge Q(z) is collected over 60 seconds and is corrected for ambient temperature and pressure using the correction factor C TP . Additionally, the reading used these correction factors: electrometer calibration K Elec , beam quality K Q , chamber conversion K Ka→Kw , and ionization chamber calibration N k . The vendor and other investigators have endorsed the use of the T30 spectrum for the INTRABEAM spectrum; thus, the beam quality correction factor was set to unity (K Q = 1). 7,9 The chamber conversion factor converts airkerma measurements to dose in water for the chamber in a T30 beam and is reported by the manufacturer on the chamber calibration certificate. 8 For the PTW model 34013 ionization chamber, the vendor defines the effective point of measurement to be inside of the chamber's entrance foil. 15 Thus, when the chamber is inserted into the waterproof holder, a measurement setup offset Δz using Equation (2) must be calculated.  (4) where the coverage factor of an expected distribution is k, and k = 1 is one standard deviation.
The standard deviation from three chamber measurements, σ rep , the dose uncertainty due to a chamber positioning error, σ pos , and the cumulative uncertainty of the chamber calibration; σ cal , are needed. The σ pos is estimated by measuring the dose rate at ±0.01cm depths and taking the average of the deviation. The uncertainty stated on the chamber calibration certificate, according to guidelines of BIPM, is σ cal = 2% (k = 1).

2.E | Physical dose delivery accuracy
The AAPM Task

2.F | Linear-quadratic model
To evaluate the impact that revised dosimetry will have on the biological dose, we present the percent change in EQD2, which is a function of the fractionation sensitivity parameter α/β and the generalized Lea-Catcheside time factor g. The g factor is introduced to consider mono-exponential repair kinetics during treatment delivery.
In the case of a constant dose rate, the generalized Lea-Catcheside equation can be expressed as Equation (5).
The treatment duration t ranges from 0. 12  Other investigators have reported shorter The calculation of EQD2 using the linear-quadratic (LQ) model 21 is expressed in Equation (6).

| RESULTS
For measurement uncertainty, Table 1 shows the standard deviation from three chamber measurements, σ rep is < 0.25% and σ V4:0 k¼1 ð Þ is < 3.5%, for the 1.0-, 2.0-, and 3-cm investigated depths. Table 2 indicates a ≤ 2.48% difference between the V4.0 calibration DDC and the PTW model 34013 ion chamber measurement. Our results are within the measurement uncertainity budget (<3.5%) shown in Table 1.   ical applicators, an important concern is that the < 20% dose deviation threshold for medical events was exceeded. The difference in % is within measurement uncertainty σ V4.0 (k = 1) as presented in Table 1.     In respect to the Al filter within ≤ 3.0-cm diameters, spherical applicator variances are shown. Table 3 shows the decreased TAR-GIT and V4.0 dose rates when the diameter of the spherical applicator increased except for the 3.0-to 3.5-cm diameter spherical applicators. Table 4 shows the increased delivery time when the diameter of the spherical applicators increased except from the 3.0to 3.5-cm diameter spherical applicators. Within Table 6, expected and delivered doses increased as the diameter of the spherical applicator increased except for the 3.0-to 4.0-cm diameter spherical applicators.

| DISCUSSION
The results of this study demonstrate that the dose rate, delivered dose, and EQD2 values are higher than previously expected with the maximum differences observed for the 1.5-cm diameter spherical applicator. Because more considerable differences in dose rate are shown at depths proximal to the XRS and a smaller spherical applicator surface is more proximal to the XRS, there is an increased concern of dose accuracy with the use of smaller spherical applicator sizes. Among patients in the North American TARGIT trial, the most used spherical applicator diameters were 3.5 cm (23%) and 4 cm (35%). 24 The 1.5-cm diameter spherical applicator has a volume of 1.8 cm 3 and is used for in situ ductal carcinoma cases, which tend to be less often treated in our clinic. Thus, most centers do not purchase the 1.5-and 2.0-cm diameter spherical applicators.   32 Retrospective dosimetric reviews showed that the 60-Gy prescription was equivalent to a 56-Gy prescription with heterogeneity corrections. 32 In response, RTOG 0613 adopted a 54-Gy prescription and mandated heterogeneity corrections. 33  The mean α/β = 1.2 Gy was used to calculate the data.
T A B L E 5 Display of dose data at the applicator surface, when 20-Gy is prescribed using the TARGIT DDC.

CONF LICT OF I NTEREST
The authors declare no conflict of interest.