Radiation treatment planning and delivery strategies for a pregnant brain tumor patient

Abstract The management of a pregnant patient in radiation oncology is an infrequent event requiring careful consideration by both the physician and physicist. The aim of this manuscript was to highlight treatment planning techniques and detail measurements of fetal dose for a pregnant patient recently requiring treatment for a brain cancer. A 27‐year‐old woman was treated during gestational weeks 19–25 for a resected grade 3 astrocytoma to 50.4 Gy in 28 fractions, followed by an additional 9 Gy boost in five fractions. Four potential plans were developed for the patient: a 6 MV 3D‐conformal treatment plan with enhanced dynamic wedges, a 6 MV step‐and‐shoot (SnS) intensity‐modulated radiation therapy (IMRT) plan, an unflattened 6 MV SnS IMRT plan, and an Accuray TomoTherapy HDA helical IMRT treatment plan. All treatment plans used strategies to reduce peripheral dose. Fetal dose was estimated for each treatment plan using available literature references, and measurements were made using thermoluminescent dosimeters (TLDs) and an ionization chamber with an anthropomorphic phantom. TLD measurements from a full‐course radiation delivery ranged from 1.0 to 1.6 cGy for the 3D‐conformal treatment plan, from 1.0 to 1.5 cGy for the 6 MV SnS IMRT plan, from 0.6 to 1.0 cGy for the unflattened 6 MV SnS IMRT plan, and from 1.9 to 2.6 cGy for the TomoTherapy treatment plan. The unflattened 6 MV SnS IMRT treatment plan was selected for treatment for this particular patient, though the fetal doses from all treatment plans were deemed acceptable. The cumulative dose to the patient's unshielded fetus is estimated to be 1.0 cGy at most. The planning technique and distance between the treatment target and fetus both contributed to this relatively low fetal dose. Relevant treatment planning strategies and treatment delivery considerations are discussed to aid radiation oncologists and medical physicists in the management of pregnant patients.


| INTRODUCTION
Patients requiring radiation therapy are seldom simultaneously pregnant. However, when both conditions apply, unique considerations are required from the radiation oncologist and the medical physicist.
Especially at doses exceeding 10 cGy, the deterministic effects of ionizing radiation on the developing fetus are moderately understood, and linear extrapolations of stochastic risk estimates are commonplace. 1,2 Radiation therapy can play a net-beneficial role in the management of a pregnant patient, but depending on the treatment site, special treatment planning techniques to reduce peripheral dose and/or fetal radiation shields may be necessary.
While breast cancer and hematologic malignancies make up the preponderance of cancers seen in a pregnant population, other tumor types are found with some frequency, including brain tumors. 3 Common brain malignancies (e.g., gliomas) found in a patient population of child-bearing age are often treated with shaped radiation fields or arcs that enter the patient's head from many angles, including so-called "vertex" beams. It may be possible to achieve a clinically acceptable plan while substantially reducing peripheral dose by modifying these standard treatment planning strategies. Numerous reports have detailed planning strategies to reduce peripheral dose. [4][5][6][7][8][9][10] While IMRT is a common choice for intracranial treatments, IMRT often results in higher peripheral dose than 2D-or 3D-conformal treatment techniques. 5,10 The purpose of this manuscript was to detail the special considerations for a pregnant brain cancer patient recently treated in our clinic, including treatment plan design for both a Varian TrueBeam system and an Accuray TomoTherapy HDA system, peripheral dose estimation and measurement, and other patient management strategies. While prior reports have provided estimates of peripheral dose for various combinations of beam energy and geometry, these reports often apply to a prior generation of treatment delivery system. 1,[4][5][6][10][11][12][13][14] The present manuscript details fetal dose measurements on current-generation treatment delivery systems and summarizes the anticipated risks to the patient's fetus using published guidance. Finally, we aim to briefly summarize some of the relevant literature on fetal dose in radiation therapy.

2.A | Patient details
The patient detailed in this report is a 27-year-old pregnant female.
She was simulated for treatment to her grade 3 astrocytoma resection cavity during gestational week 17. The patient was prescribed 50.4 Gy in 28 fractions to a primary target volume to be followed sequentially by a 9 Gy boost in five fractions to a smaller volume.
Medical images with delineated target volumes are shown in Fig. 1.
Standard departmental brain planning constraints were ordered for this patient, including D(0.03 mL) < 54 Gy for the brainstem, optic chiasm, and optic nerves; mean dose < 35 Gy and D(0.03 mL) < 40 Gy for the cochleae; D(0.03 mL) < 7 Gy for the lenses of the eyes; and D(0.03 mL) < 45 Gy for the spinal cord. At least 95% of the target volume was covered with 99% of the prescription dose in each treatment plan, and target hotspots were maintained less than 110%. | 369 symphysis were measured from a fixed radio-opaque marker placed on her chin. Palpation of the uterine fundus was not achieved in our department; instead, a brief consultation with diagnostic radiology immediately prior to her CT simulation measured her uterine fundus to be 4 cm inferior to her umbilicus using a portable ultrasound unit.
The total distance from each point of interest to the target volume was later determined by measuring the distance between the radioopaque marker and the segmented target volume in the simulation CT image. Finally, we assumed superior progression of the patient's uterine fundus at a rate of 1 cm/week and that the patient would be treated during gestational weeks 19-25. 1 Table 1 shows the distances from the target volumes to the points of interest.
Our institution's Health Insurance Portability and Accountability Act (HIPAA) Privacy Officer has reviewed this manuscript to ensure compliance with our institution's standards for protected health information. Institutional Review Board review was not required.

2.B | Treatment plans
In an attempt to use available department resources to reduce fetal dose, four potential treatment plans were developed for the patient: All treatment plans used specific planning techniques to reduce peripheral dose. The TrueBeam treatment plans used collimator rotations of 90°to place the distal x jaws in the patient superior-inferior direction and avoided the use of physical wedges; EDW fields may increase peripheral dose by 10%-20% in very close proximity to the treatment field, compared to 200%-400% increases for physical wedges. 4,5,7,8 Tertiary MLC collimation was used for all TrueBeam treatment plans. The TrueBeam plans had an isocenter placed as far cranially as possible to maximize the separation between the treatment head and the fetus 9 ; additionally, couch kicks were avoided to maximize separation between the treatment head and the fetus and to avoid beam divergence toward the fetus. Beam energy was limited to 6 MV to reduce scatter, head leakage, and neutron contamination, and the flattening filter-free beam was investigated to assess head leakage reductions from the removal of the flattening filter from the beamline. 5

2.C | Estimates of fetal dose
Fetal dose was estimated using the prescribed doses given previously, the distances from treatment volumes to points of interest (Table 1)

2.D | Measurement of fetal dose
Fetal dose was measured at each of the four points of interest identified in Table 1  While out-of-field dose has minimal dependence on depth, the superficial dose is markedly higher than the rest of the depth-dose curve; the bolus was used to position the TLDs beyond this superficial region. 11 In addition to the TLDs, a Farmer-type ionization cham-

2.E | Patient imaging
The patient was imaged for treatment planning using our department's CT simulator. A 0.5 mm lead-equivalent apron was placed around the patient's abdomen and pelvis during her CT scan; this type of apron may reduce the CT dose to the fetus by 90% or more. 16  For image guidance during patient treatment, the intended strategy depends on the treatment. For the TrueBeam plans discussed above, orthogonal planar 2D kV-kV imaging would be entirely sufficient to align the cranium with five of the available six degrees-offreedom. If cone-beam CT (CBCT) were required, it could be performed with minimal fetal dose on a modern TrueBeam system.
Scaling organ doses previously published for a prior generation of Varian CBCT system by the CTDI ratios between the older On-Board Imager (OBI) CBCT system and the TrueBeam CBCT system, one could estimate up to 0.0006 cGy fetal dose per "head" CBCT and up to 0.005 cGy fetal dose per "thorax" CBCT (using kidneys as a conservative surrogate for the fetus). [19][20][21] TomoTherapy image guidance would be performed with the "coarse" MVCT protocol with 3 mm slice thickness to minimize the MU required for MVCT imaging.

| RESULTS
The patient started treatment during gestational week 19 and fin- Inverse-square scaling may be appropriate given that the preponderance of peripheral dose for this study originates as leakage from the treatment head itself. 22 The measured doses for the four plans at the four points of interest are also given in

| DISCUSSION
This study reports on a brain tumor patient treated recently in our clinic to a total dose of 59.4 Gy in 33 fractions during her second trimester of pregnancy. Four candidate treatment plans were created using a variety of planning strategies intended to reduce fetal dose.
Estimates of fetal dose for each plan were made using published literature, and measurements of fetal dose were made for each plan using a modified anthropomorphic phantom. Measurements and estimates were performed at points of interest that bracketed the potential locations of the fetus during the duration of the patient's treatment. The treatment plan with the lowest measured fetal dose was selected for patient treatment.
The estimated doses to the points of interest were adjusted using the MU-to-cGy ratio for the TrueBeam treatment plans, as well as using the inverse-square law when peripheral dose estimates were not available at the distances given in Table 1 6 MV-FFF for a 3D brain plan resulted in a 20% reduction in unshielded fetal dose. 5 As a note of caution, using 6 MV-FFF for T A B L E 2 Estimated and measured fetal doses to the points of interest in Table 1 from the four candidate treatment plans. Fetal dose estimates were taken from Mutic and Klein, 4 Owrangi et al., 5 and Lissner et al. 14 The Lissner and Owrangi dose estimates were inverse-square corrected, and the Mutic/Owrangi estimates were increased by the ratio of MU-to-cGy for plans 1-3. Estimated uncertainty was 5% for TLD measurements and 7% for ion chamber measurements.