Extraneous dose is the unavoidable low dose radiation which the patient receives outside the planned target volume. In all radiation treatment modalities, including IMRT, normal tissues receive low dose radiation from head leakage, patient internal scatter, and collimator scatter sources (Fig. 22.3). The main disadvantage of IMRT is that a larger volume of unintended normal tissues are exposed to low dose radiation as a result of higher monitor units with complex multi-field plans and increased treatment times and greater neutron contamination for higher energy treatments [18]. Some authors argue that the extraneous low dose radiation may contribute to a higher rate of secondary malignancy(s). While the estimation of radiation-induced secondary malignancy(s) is an area of active research, radiation biologists have modeled risks for second malignancy(s) in the low dose region to be approximately “8 % per Gy probability for a fatal cancer to develop due to radiation above spontaneous incidence. An estimated 0.75–1 % of surviving patients would be expected to develop a second malignancy as a consequence of IMRT, approximately double the incidence of secondary malignancies due to conventional therapy” [19–22].

In addition to second malignancy(s), radiation-induced sequelae to radiosensitive organs outside the radiation field need to be considered. Organs/tissues which may be highly susceptible to extraneous low dose radiation include gonads, breast tissue, thyroid, and lens (infertility or sterility, cessation or abnormal growth, hypothyroidism, cataracts) [23, 24]. Irreversible radiation-induced injury to the gonads can significantly alter quality of life for childhood cancer survivors who desire an offspring later in life. Since the gonads are one of the most radiosensitive organs, even exposure to low dose radiation may inadvertently contribute to infertility or sterility [25]. Thyroid and lens are also very radiosensitive structures which can be affected by very low dose radiation. Breast tissue is highly susceptible to the mutagenic effects of ionizing radiation in young females. The risk for breast cancer in Hodgkin’s lymphoma survivors who receive chest irradiation remains high even for patients treated with conformal techniques such as IMRT. Recent reports show that the majority of female childhood cancer survivors who receive low dose (median 14 Gy) radiation therapy to large volumes (e.g., whole-lung irradiation) have a significantly elevated risk for developing breast cancer [26].

While studies show that differences between IMRT and 3DCRT regarding peripheral extraneous doses received by gonads, thyroid, breast tissue, and lens in children and young adults are minimal [27], they should always be considered, calculated, and/or directly measured for IMRT plans. Dose distributions therefore may influence clinical decisions where pediatric radiation oncologists may consider 3DCRT versus IMRT to protect sensitive structures from the low dose irradiation. However, it is important to know that the overall absolute extra-target peripheral dose for the two modalities is influenced by different “competing” dominant effects (Fig. 22.3). For 3DCRT patient, internal scatter is larger resulting in higher out-of-field doses closer to the target region, while with IMRT, there is an increased head leakage corresponding to higher peripheral doses far from the target region [28]. Therefore, the increased monitor units delivered with IMRT should not be treated as a single, accurate indicator of increased out-of-field peripheral irradiation and subsequent low dose exposure of normal tissues [28].

Given the observed benefits of organ-/tissue-sparing capabilities of IMRT in the high-dose region, several strategies have been developed and demonstrated to minimize the unintended out-of-field dose from IMRT delivery. Using coplanar IMRT beams may be one technique to minimize the internal scatter dose. Kan et al. reported on five cases with different pediatric malignancies in the head and neck planned with both coplanar and noncoplanar IMRT techniques. It was observed that peripheral doses were 1.8–2.5 times higher while using the noncoplanar beams [29]. Another technique to reduce out-of-field dose is the removal of the flattening filter, which removes a neutron contamination source associated with high-energy photons interacting in the filter, reduces head leakage, and reduces the delivery time. Cashmore et al. tested this hypothesis with a linear accelerator outfitted with conventional (flattened) and flattening-filter-free modes. IMRT treatment plans for pediatric intracranial treatments were delivered to phantoms using both approaches. Measurements indicated a 23–70 % reduction in peripheral doses (from thyroid to gonads) using the flattening-filter-free modes [30].

The reduction in beam-on time (number of monitor units) is also a method to minimize out-of-field peripheral doses. Besides significantly increasing target coverage and reducing the delivery time, VMAT delivery of IMRT plans has been shown to generally match or even reduce the treatment monitor units compared to conventional static field IMRT plans [31]. In complex pediatric pelvic cases, VMAT reduced the treatment time by 78 % and monitor units by 25 % compared with standard IMRT as reported by Matuszak et al. [32]. The reduction in out-of-field peripheral dose associated with the delivery and reduction in monitor units was as high as threefold for thyroid during pelvic irradiation [31]. There is a variety of commercially available treatment planning systems, each offering optimization solutions that can be employed in reducing extraneous dose in pediatric patients. Attalla et al. reviewed current planning systems offering IMRT optimization. In this study, IMRT plans using step-and-shoot were designed for pediatric head and neck and CNS cases necessitating simultaneous integrated boosts. The authors observed that while three different commercial treatment planning systems were used to obtain optimized IMRT plans achieving the same clinical objectives, using the same energy, number, and direction of beams, the resulting plan quality was not comparable. They found one system was superior producing more efficient plans, fewer segments and MUs, shorter treatment delivery times, and better conformality [33]. Dose painting (DP), simultaneous integrated boost (SIB), and simultaneous modulated accelerated radiotherapy (SMART) are planning techniques which allow highly customized, highly conformal dose distributions while treating multiple targets to different dose levels and sparing more normal tissue [34, 35].

Due to its complex resulting spatial distributions of dose deposition and high gradients between targets and adjacent normal tissues, IGRT is necessary when using IMRT. Even with anesthesia minimizing the risk for movement during treatment, the precision in executing the treatment plan relies on accurate alignment with daily in-room verification using kV imaging. The use of three-dimensional alignment using cone beam computed tomography (CBCT) adds additional accuracy of setup alignment, particularly in the head and neck region. Additionally, it aids in the decision as to whether replanning is necessary for tumors regressing during the course of treatment or for anatomical changes during the course of radiation such as rapid weight loss. The benefit from frequent CBCT portal imaging must be carefully weighed against the risk for exposure to nontherapeutic ionizing radiation. Daily IGRT imaging doses should be recorded for the cumulative exposure above the prescribed therapeutic dose [36]. The dose contribution from imaging is generally less than 2–3 % of the prescribed dose and is usually neglected from total dose summations [37]. However, for pediatric patients, sensitive structures such as the testes or ovaries may be affected if daily kV planar imaging or CBCT was to be used. For example, testicular doses from kV imaging can be 3–4 times higher than the actual incidental dose from pelvic irradiation treatments. Reducing the imaging field of view to exclude the testes may significantly decrease dose to this region [38, 39].

While refinements in delivery methods, increasing positional accuracy, and decreasing delivery time are critical in minimizing extraneous radiation dose exposure in pediatric patients, the physics of photons’ interaction with tissue does not allow for complete avoidance of normal tissue exposure to low dose radiation. The pediatric oncology community has been investigating intensity modulated proton therapy (IMPT) as the step beyond IMRT in radiotherapy [40]. Many proponents of proton therapy to treat childhood cancer refer to the desirable beam characteristics which minimizes the out-of-field dose (see Fig. 22.3). Protons, unlike photons, deposit most of their energy at the distal end of their range (Bragg peak). This implies practically nonexistent exit dose, while the entrance dose is lower compared to photon attenuation in tissue [41].

There are two main delivery methods in proton therapy: passive scattering and spot scanning. Passive scattering employs beam scatterers to spread out the input beam laterally, metal apertures to collimate the beam, and range modifying devices to spread out the Bragg peak. In principle, passive scattering employs both beam energy and intensity modulation. However, IMPT refers to the second delivery method. Spot scanning, also referred to as pencil beam scanning or modulated scanning, delivers the treatment layer by layer in raster format scanning a pencil beam using powerful magnets. Depth is changed by switching energy, hence changing the range of the protons. Modern delivery systems are increasingly efficient, allowing for very quick delivery of dose even for large target volumes [41].

The primary advantage that high-energy proton therapy has over photon therapy is reducing normal tissue dose [41]. Several studies have modeled the relative risk associated with normal tissue irradiation for both proton treatments and current state-of-the-art IMRT. Athar et al. reported that the potential organ-specific second cancer lifetime attributable risks would be from unintended internal scatter or leakage out-of-field low dose irradiation from 6 MV IMRT and passive scattering proton therapy. The modeling study included data from patients ranging in age from 9 months to 14 years old and one adult and two treatment sites (brain and mid-spine). The lifetime attributable risk for developing a thyroid cancer after treatment to a brain lesion in a 4-year-old patient was estimated at 1.1 % for IMRT versus 0.3 % for proton therapy. The lifetime attributable risk for developing a bladder cancer after treatment to a mid-spine field was estimated to be 0.2 % with IMRT and 0.02 % with proton therapy, suggesting a distinct advantage to proton beam especially with regard to organs at risk further away from the field [42]. Lifetime attributable risks were also modeled to assess the risk for developing radiation-induced tumors within the path of the beam. Using whole-body phantoms for a 4- and 14-year-old child, plans were generated for optic glioma and vertebral body Ewing’s sarcoma. The lifetime attributable risks were modeled for these cases, and the results showed that the risks associated with proton therapy were lower by a factor of 2–10 in the case of protons [43]. In both cases, long-term follow-up is needed to confirm the premise that proton beam therapy is associated with fewer radiation-induced secondary malignancy(s) [44].

Greco et al. reviewed the current trends in proton therapy for children and concluded that IMPT has the potential to “yield superior dose distributions to photon IMRT, with the added advantage of a significant reduction in the volume of healthy normal tissues exposed to low-to-medium doses” [45]. Lomax et al. emphasized the versatility of scanning beam IMPT, where the individual Bragg peaks can be delivered from any field and can be distributed in 3D throughout the target volume, providing an increasing amount of degrees of freedom for designing dose distributions when compared to IMRT or conventional proton therapy [46].

There are problems to be solved with IMPT. The protons relative biological effectiveness (RBE) is spatially variable across the energy deposition curve and maximized at the Bragg peak [47]. Furthermore, the uncertainty in the range of protons in tissue is affected by imaging, patient setup, beam delivery, and dose calculation techniques [48]. High-energy charged particles such as protons yield a significant neutron spectrum in their interaction with the beam line components and the patients’ body, increasing the out-of-field dose [49, 50]. A major hurdle to overcome is that proton facilities are very costly to build and operate, limiting the feasibility of using this modality in many places around the world.