A dosimetric comparison of proton and photon therapy in unresectable cancers of the head of pancreas

Transcription

1 A dosimetric comparison of proton and photon therapy in unresectable cancers of the head of pancreas Reid F. Thompson University of Pennsylvania, Philadelphia, Pennsylvania 1914 Sonal U. Mayekar Thomas Jefferson University, Philadelphia, Pennsylvania 1917 Huifang Zhai and Stefan Both University of Pennsylvania, Philadelphia, Pennsylvania 1914 Smith Apisarnthanarax University of Washington, Seattle, Washington 9819 James M. Metz, John P. Plastaras, and Edgar Ben-Josef a) University of Pennsylvania, Philadelphia, Pennsylvania 1914 (Received 2 March 214; revised 23 June 214; accepted for publication 24 June 214; published 22 July 214) Purpose: Uncontrolled local growth is the cause of death in 3% of patients with unresectable pancreatic cancers. The addition of standard-dose radiotherapy to gemcitabine has been shown to confer a modest survival benefit in this population. Radiation dose escalation with three-dimensional planning is not feasible, but high-dose intensity-modulated radiation therapy () has been shown to improve local control. Still, dose-escalation remains limited by gastrointestinal toxicity. In this study, the authors investigate the potential use of double scattering () and pencil beam scanning () proton therapy in limiting dose to critical organs at risk. Methods: The authors compared,, and plans in 13 patients with unresectable cancer of the pancreatic head, paying particular attention to duodenum, small intestine, stomach, liver, kidney, and cord constraints in addition to target volume coverage. All plans were calculated to 55 cgy in 25 fractions with equivalent constraints and normalized to prescription dose. All statistics were by two-tailed paired t-test. Results: Both and decreased stomach, duodenum, and small bowel dose in low-dose regions compared to (p <.1). However, protons yielded increased doses in the mid to high dose regions (e.g., and Gy for duodenum using and, respectively; p <.5). Protons also increased generalized equivalent uniform dose to duodenum and stomach, however these differences were small (<5% and 1%, respectively; p <.1). Doses to other organs-at-risk were within institutional constraints and placed no obvious limitations on treatment planning. Conclusions: Proton therapy does not appear to reduce OAR volumes receiving high dose. Protons are able to reduce the treated volume receiving low-intermediate doses, however the clinical significance of this remains to be determined in future investigations. 214 American Association of Physicists in Medicine. [ Key words: pancreas, unresectable, protons,, 1. INTRODUCTION Pancreatic cancer is an aggressive malignancy with only approximately 1% of patients found to have curable, resectable disease at the time of diagnosis. More often, local tumor invasion into adjacent vessels (e.g., celiac and superior mesenteric arteries) precludes complete resection, with eventual local progression and distant metastasis. While dissemination of disease is certainly life-limiting, local growth and invasion is the cause of death in approximately one third of patients with unresectable pancreatic primaries. 1 Chemoradiotherapy is a common treatment for locally advanced disease, and may provide good palliation and even survival advantage. Indeed, the combination of gemcitabine and standard-dose radiotherapy has been shown to confer a modest survival benefit in this setting. 2 Unfortunately, radiation regimens in clinical use are generally limited to inadequate tumor doses (e.g., 5 54 Gy in Gy daily fractions) due to the sensitivity of neighboring organs-at-risk (OARs) such as the duodenum and the stomach. Radiation dose escalation has been attempted using three-dimensional conformal techniques to doses of 7 72 Gy, however this approach was complicated by severe gastrointestinal toxicity. 3 Dose escalation with more advanced treatment techniques such as intensity-modulated radiation therapy () offers improved local control and survival; however, dose also remains constrained by grade 3 4 gastrointestinal toxicities Med. Phys. 41 (8), August /214/41(8)/81711/1/$ Am. Assoc. Phys. Med

2 Thompson et al.: Proton dosimetry in pancreatic cancer Proton therapy offers theoretical advantages over photonbased techniques and may be of benefit in radiation dose escalation. In particular, the rapid falloff of distal dose in proton therapy allows minimization of integral dose to surrounding OARs. While double-scattering () proton therapy does not allow for proximal conformality, pencil beam scanning () proton therapy can achieve improved dose conformality, further sparing dose to nearby OARs such as bowel, stomach, and liver. We therefore hypothesized that for similar target dose, proton therapy could reduce dose to OARs. We further hypothesized that would enable greater dose reduction to OARs when compared with the technique. In this study, we compare the in silico dosimetry of as well as and proton therapy plans, focusing particular attention on doses to neighboring OARs. 2. METHO 2.A. Study population With approval from the University of Pennsylvania Institutional Review Board, we identified 13 patients with unresectable cancer of the pancreatic head who were treated with definitive chemoradiation from 28 to 211 (see Table I for detailed patient characteristics). 2.B. Target volume and OAR definition Gross tumor volume (GTV) was contoured on IV contrast enhanced CT simulation scans. Clinically apparent nodes were included along with gross disease, however elective nodal regions were not included in target volumes. 5 GTV to clinical target volume (CTV) and CTV to planning target volume (PTV) uniform expansions of.5 cm were based on the previously published phase I/II dose escalation trial. 4 For the purpose of this study, it was assumed that all patients would be treated using breath-hold technique, discounting motion in order to assess maximal potential benefit. Relevant OARs including duodenum, small intestine, stomach, liver, kidney, and spinal cord were all contoured according to standard tech- nique. Note that small intestine was contoured as individual loops of bowel. 2.C. Treatment planning All,, and plans were generated by a single dosimetrist (Z.H.) for each of 13 different patients with unresectable cancer of the pancreatic head. fields employed seven coplanar 6 MV photon beams with even distribution of gantry angles at 52 offsets. and plans included three noncoplanar MeV proton fields (right superior posterior oblique, posterior, and posterior superior oblique at gantry angles 215, 17, and 16, respectively, incorporating couch kicks as needed). This noncoplanar arrangement was based on prior optimization of beam angles and was chosen to mitigate kidney dose, spare skin, and minimize duodenal dose contribution from beam penumbra. 6, 7 An example beam arrangement is shown in Fig. 1. Beam angles and weights varied on a per patient basis. Note that this target prescription dose and OAR constraints for planning were based on guidelines from a prior prospective phase 2 doseescalation trial 4 as follows: 99% of CTV receiving at least 98% of prescription dose; 98% of PTV receiving at least 95% of prescription dose; maximum kidney dose bilaterally 2 Gy, with less than 1% of their volume receiving 18 Gy; mean liver dose 3 Gy (and as low as reasonably achievable); maximum stomach (and small bowel) dose 54 Gy, TABLE I. Patient characteristics. AGE Gender CTV (cc) PTV (cc) Stent Stage 56 M N T4NM 59 F N T4NM 6 F Y T4NM 45 M Y T4N1M1 69 F Y T3N1M 65 M Y T4N1M 48 M N T4N1M 73 F Y T3N1M 62 F N T3N1M 67 M N T3N1M 79 M Y T4N1M 69 F N T4N1M 61 M N T4N1M FIG. 1. Arrangement of three noncoplanar proton fields for a single patient. Three-beam arrangement (schematic projections defined by yellow lines) and dose distribution (color wash overlay) are shown for a single patient on axial (a) and sagittal (b) planes. The noncoplanar arrangement is shown as a three-dimensional reconstruction (c) with right superior posterior oblique, posterior, and posterior superior oblique fields at gantry angles 215, 17, and 16, respectively. Medical Physics, Vol. 41, No. 8, August 214

3 Thompson et al.: Proton dosimetry in pancreatic cancer with 2% of its volume exceeding 5 Gy and 25% volume exceeding 45 Gy; spinal cord maximum dose 45 Gy; maximum duodenal dose 55 Gy (prescription dose), with 33% of its volume exceeding 45 Gy. For and plans, target prescription and OAR constraints were identical to. Distal and proximal margins were applied to correct for range uncertainties of 3.5%+3 mm and +1 mm, respectively, as previously described. 8, 9 Appropriate aperture margins and smearing were also applied to multileaf collimator and compensator shapes for. For optimization of plans specifically, a pencil beam scanning target volume (TV) was created using an asymmetric 8 mm expansion (5 mm posteriorly) and single-field uniform dose. All plans were calculated to a prescription dose of 55 cgy in 25 daily 22 cgy fractions, 4 and were normalized to 1% of the CTV mean dose. Calculated isoeffective proton doses all assumed a relative biological effectiveness of D. Data analysis We compared,, and plans in 13 patients with unresectable cancer of the pancreatic head treated to 55 cgy in 25 fractions, paying particular attention to duodenum, small intestine, stomach, liver, kidney, and spinal cord constraints in addition to target volume coverage. Dosevolume histogram (DVH) data were exported directly from Eclipse and analyzed using the RadOnc package (1..3) for R (3..2). 11 We assessed standard dosimetric parameters as well as integral biological equivalent doses (BED; calculated using the linear quadratic model: D [1+ d/(α/β)], where D is total dose, d is dose per fraction, and assuming α/β ratio of 3 for all OAR) to model biological effect for each OAR. Generalized equivalent uniform dose (geud) models were also applied for duodenum, stomach, and small bowel to summarize the respective DVHs as single biologically weighted values (evaluated using tissue-specific parameter a = 6 and a = 1 as in prior work). 12 All DVH comparisons were by two-tailed paired t-test using α =.5 (unless otherwise indicated) and applied to individual points across the entire dose range. 3. RESULTS 3.A. Target volume coverage CTVs ranged from 6 to 225 cc (mean 122 cc), with corresponding PTVs of cc (mean 25 cc). Target volume coverage for all plans met standard institutional constraints (e.g., V95% > 95%) and was clinically comparable among both proton and photon plans (Fig. 2). However, proton therapy offered incrementally better dosimetric coverage compared to (V95% of 98.9% vs 97.5% for protons and, respectively; p =.2, unpaired t-test). 3.B. Dose to organs-at-risk Dose to adjacent OARs differed substantially between and proton plans. Irrespective of technique, proton therapy conferred large dosimetric benefit to the liver, with 5% reduction in mean dose in all cases compared to (p =.1 or p =.5 using paired and unpaired tests, respectively). However, mean dose was well within standard constraints in all cases and there were only minimal differences in hepatic dose above 25 cgy [Figs. 3(a) 3(c)]. CTV PTV % % FIG. 2. Target volume DVH curves for photon and proton plans. Average DVH curves are shown for each cohort of plans (e.g.,,, and ) with dose shown along the x-axis and cumulative volume receiving at least that dose along the y-axis. Two gray vertical lines in each panel denote the 95% and 1% of prescription dose lines. CTV and PTV data are segregated in two panels as labeled. Medical Physics, Vol. 41, No. 8, August 214

4 Thompson et al.: Proton dosimetry in pancreatic cancer Liver (e) Small Bowel Duodenum 1 3 (k) (l) (j) Duodenum Duodenum (i) (h) 5 4 Small Bowel (g) Small Bowel (f) Stomach (d) Stomach Stomach (c) (b) (a) Liver Liver F IG. 3. Comparisons of DVH data for liver, stomach, small bowel, and duodenum using paired,, and plans. Each panel represents a paired statistical comparison of DVH curves from two different treatment modalities for a given organ at risk. Panels are labeled according to the compared structure, with liver, stomach, small bowel, and duodenum shown in rows 1 4, respectively. Figure columns are ordered by comparison group, with vs, vs, and vs shown in columns 1 3, respectively. In all cases, radiation dose is shown along the x-axis and the cumulative volume receiving at least a given dose is plotted along the y-axis. Shaded curved regions represent the 95% confidence intervals derived from paired statistical comparisons. Each comparison also contains an upper panel plot of statistical significance, shown in inverse log-scale along the y-axis with numerical p-values resulting from paired statistical comparison between the DVH data in each group. Highlighted regions of these curves represent significant differences (i.e., p <.5) between DVH groups at a given dose. Medical Physics, Vol. 41, No. 8, August 214

5 Thompson et al.: Proton dosimetry in pancreatic cancer Similarly, proton therapy decreased stomach and small bowel dose in low-dose regions (// V2 Gy: 21.4/11.1/9.3% and 19.7/9.8/6.5% for stomach and small bowel, respectively, p <.1 or p <.5 using paired and unpaired tests). However, protons yielded incremental increase in the high dose region for stomach (e.g., V45 Gy: 3.9% vs 5.8% vs 4.5% for,, and, respectively, p <.1; no significant differences by unpaired t-test) and small bowel V45 Gy (2.4% vs 4.2% vs 2.9% for,, and, respectively, p <.5; no significant differences by unpaired t-test). Figures 3(d) 3(i) depict the comparative dosevolume relationships among protons and photons in more detail. While protons demonstrate clear dosimetric benefit below 3 Gy, is incrementally superior above 4 Gy for both stomach and small bowel. In all cases, paired comparison between and plans demonstrates a relative superiority of throughout the entire dose range [Figs. 3(f) and 3(i)]. Dose to the duodenum exhibited a similar pattern of differences between and proton plans [Figs. 3(j) 3(l)]. Specifically, protons decreased dose to the duodenum in the lowest dose region (<3. and <3.5 Gy for and, respectively; p <.5) and increased dose in the higher dose region ( and Gy for and, respectively; p <.5) when compared with. Of note however, there was also a narrow region of dose ( Gy) where demonstrated a significant decrease in duodenal dose compared to (e.g., V55.25 Gy of 5.4% vs 2.8% for and, respectively, p =.2; no significant difference by unpaired t-test). This dose difference is accounted for by the differential gradient achieved with and, shown in Fig. 4 for all patients in this cohort. Note that this phenomenon was not observed with [Fig. 3(j) and data not shown]. Doses to other organs-at-risk were within all institutional constraints and placed no obvious limitations on treatment planning. Nonetheless, proton and plans exhibited different patterns of dose for spinal cord and kidneys (Figs. 5 and 6). Overall, decreased mean kidney dose by 18% compared with (7.8 Gy vs 9.5 Gy for and, respectively, p =.1; no significant difference by unpaired t- test). Predictably, protons also decreased overall integral dose to the body, with a 32% difference compared to (p <.5 using either paired or unpaired t-tests). This was independent of proton modality, with the same 32% integral dose reduction using both and techniques [Figs. 5(g) 5(i)]. 3.C. Biologically equivalent and equivalent uniform dose analysis Due to the opposing relationships between protons and photons in the lower and higher-dose regions, we sought to further evaluate the biological relevance of these distributions in duodenum, stomach, and small bowel. We computed BEDweighted dose distributions and corresponding dosimetric parameters for each structure. and both decreased BEDweighted mean doses compared to (//: 15.7/13.6/12. Gy for stomach and 14.1/11.1/8.9 Gy for small bowel, respectively, p <.5; no significant differences using unpaired t-test). To further investigate the varying dose distributions produced in proton and photon plans, a geud model was implemented for duodenum, stomach, and small bowel. Both and increased average geud values for duodenum when compared with (//: 42.3/44.6/43.3 and 45.7/47.4/46.5 Gy using a = 6 or 1, respectively; p <.1, no significant differences using unpaired t-test). Similarly, geud values were increased for stomach with both and (//: 27.5/3.2/28.5 and 33.2/36.7/35.2 Gy using a = 6 or 1, respectively; p <.1, no significant differences using unpaired t-test). There were no statistically significant differences between geud values calculated for small bowel (//: 23.3/25.4/22.3 and 28.5/31.8/28.4 Gy using a = 6 or 1, respectively; p >.1). These differences are summarized in Fig DISCUSSION The doses to OARs observed in this study were mostly within our institutional constraints and are consistent with other published reports. For instance, the degree of hepatic and renal sparing in this study using protons is similar to previous findings. 13 Stomach and small bowel sparing in the lowdose region was also comparable to that observed in a prior dosimetric study. 14 Dose to the stomach and duodenum were well below the threshold of V 5 16 cm 3, with an anticipated 1% rate in this study of grade 2 or greater acute GI toxicity. 15 Similarly, duodenal dose was below V 55 1cm 3, irrespective of modality, predicting a low risk of grade 2 or greater duodenal toxicity. 16 Nonetheless, the current study demonstrates that proton therapy generally increases volumes receiving higher doses (significant by pairwise analysis but not by unpaired statistics). This is mostly explained by the additional protonspecific margins used to account for range uncertainty in the direction of the beam. For stomach and small bowel, these dose increases were relatively modest. When compared with, protons also increased dose to the duodenum in the high-dose regions; however, the highest-dose regions closely neighboring target volumes were paradoxically spared dose with compared to. This phenomenon was observed for all patients and was attributable largely to dose differences in the overlapping regions of duodenum and clinical target volume, with a more heterogeneous distribution of dose in and a more homogeneous and central distribution in plans (Fig. 4). In all cases, was dosimetrically superior or at least equivalent to. This was largely attributable to ability of the beam to conform the dose three-dimensionally while the beam ensures only distal dose conformality. We note several limitations to the current study. All treatment plan comparisons were performed in silico using an idealized treatment paradigm wherein target and organ motion Medical Physics, Vol. 41, No. 8, August 214

6 Thompson et al.: Proton dosimetry in pancreatic cancer FIG. 4. Distributions of target dose excess in paired comparisons between and plans for each patient. Each panel depicts an excerpt of a CT scan centered on the target volume for each patient; scan orientation is preserved and identical in all cases, with anterior orientation corresponding to the upper aspect of each image. In panel (a), absolute dose difference (-) is overlaid as a color gradient, with increasing dose from green to red as indicated on the scale. Negative dose differences (where dose exceeds dose) are present in regions of absent color overlay. In each panel, the gross tumor volume is encircled in light blue, with a surrounding target volume expansion in the same shade, and duodenum delineated in darker blue. The inverse absolute dose difference (-) is displayed in panel (b) with equivalent formatting to panel (a). were eliminated and reproducible delivery was assumed. To approximate this condition in practice, patients with locally advanced pancreatic cancer are treated at breath-hold. However, even under the best of circumstances, actual treatment delivery is associated with a small degree of intra- and interfraction motion. 17 The robustness of,, and in pancreatic dose escalation is a key question and will be addressed in future work. Proton dosimetry may be further improved using existent or forthcoming techniques that are not addressed in the current study. For example, collimation can improve lateral conformality 18 and decreased spot sizes may achieve better control of dose deposition. These approaches have the potential to render protons a more appealing option for pancreatic dose escalation, however they must remain the subject of future investigation. Medical Physics, Vol. 41, No. 8, August 214

7 Thompson et al.: Proton dosimetry in pancreatic cancer Spinal Cord Whole Body (i) (h) (g) Whole Body (f) Whole Body (e) (d) (c) (b) (a) Spinal Cord Spinal Cord F IG. 5. Comparisons of DVH data for spinal cord, kidney (total), and whole body using paired,, and plans. Each panel represents a paired statistical comparison of DVH curves from two different treatment modalities for a given structure at risk. Panels are labeled according to the compared structure, with spinal cord, kidneys, and body shown in rows 1 3, respectively. Figure columns are ordered by comparison group, with vs, vs, and vs shown in columns 1 3, respectively. In all cases, radiation dose is shown along the x-axis and the cumulative volume receiving at least a given dose is plotted along the y-axis. Shaded curved regions represent the 95% confidence intervals derived from paired statistical comparisons. Each comparison also contains an upper panel plot of statistical significance, shown in inverse log-scale along the y-axis with numerical p-values resulting from paired statistical comparison between the DVH data in each group. Highlighted regions of these curves represent significant differences (i.e., p <.5) between DVH groups at a given dose. The current study is confined to a dosimetric analysis of multiple treatment cohorts. It does not incorporate direct assessment of dose-limiting toxicities or address prospective clinical outcomes. Instead, clinical interpretation of the data herein is guided by prior data, which emphasize the importance of higher duodenal doses in determining subsequent Medical Physics, Vol. 41, No. 8, August 214 toxicity.15, 16, 19 This is borne out in a recent normal tissue complication probability (NTCP) model predicting severe bowel toxicities (e.g., duodenal perforation), derived from multiple dose fractionation schemes.2 While duodenal ulceration and bleeding are feared complications of dose escalation in pancreas cancer, treatment

8 Thompson et al.: Proton dosimetry in pancreatic cancer Kidney (Left) (i) (h) (g) (f) (e) Kidney (Right) (d) 4 Kidney (Right) Kidney (Right) (c) 8 8 (b) (a) Kidney (Left) Kidney (Left) F IG. 6. Comparisons of DVH data for individual and combined kidneys using paired,, and plans. Each panel represents a paired statistical comparison of DVH curves from two different treatment modalities for a given structure at risk. Panels are labeled according to the compared structure, with left kidney, right kidney, and bilateral kidneys shown in rows 1 3, respectively. Figure columns are ordered by comparison group, with vs, vs, and vs shown in columns 1 3, respectively. In all cases, radiation dose is shown along the x-axis and the cumulative volume receiving at least a given dose is plotted along the y-axis. Shaded curved regions represent the distribution of component DVHs within each group about the mean DVH for that group. Each comparison also contains an upper panel plot of statistical significance, shown in inverse log-scale along the y-axis with numerical p-values resulting from paired statistical comparison between the DVH data in each group. Highlighted regions of these curves represent significant differences (i.e., p <.5) between DVH groups at a given dose. is more often constrained by severe nausea and vomiting (approximately 6% of dose-limiting toxicities).4 It is unclear which parts of the dose spectrum are most relevant for these toxicities, however, there is some evidence to suggest that the volume receiving low-intermediate dose (e.g., V153 Gy) may be predictive of acute nausea, vomiting, and Medical Physics, Vol. 41, No. 8, August 214 diarrhea.15, 21 Also, proton therapy (with comparable beam arrangements) has been shown to reduce dose to stomach and small bowel with consequent reduction in grade 2 acute gastrointestinal toxicities.22 In summary, protons appear to increase OAR volumes receiving high dose and reduce OAR volumes receiving low to

9 Thompson et al.: Proton dosimetry in pancreatic cancer (a) Duodenum (b) Stomach (c) Small Bowel geud (cgy) geud (cgy) geud (cgy) a=6 a=1 a=6 a=1 a=6 a=1 FIG. 7. Linked distributions of calculated geud values for duodenum, stomach, and small bowel using,, and techniques. Distributions of geud values are shown along the y-axis, with labels along the x-axis corresponding to treatment modality (,, and ) and two different values for tissue parameter a (a = 6 or 1). Each geud value is linked among its corresponding values for different treatment modalities on a per-patient basis. Data are shown for three separate organs at risk: duodenuma (a), stomach (b), and small bowel (c). intermediate dose. As current NTCP models suggest that rates of severe bowel toxicity are dependent on volumes at the high dose range of the DVH spectrum, protons do not seem to hold clear advantage over for the purpose of dose escalation. However, the low-intermediate dose decreases observed with proton therapy have the potential to reduce dose-limiting nausea and vomiting. Given the limitations of current NTCP modeling, the relative clinical importance of these phenomena may be best resolved by a clinical trial. 5. CONCLUSIONS Dose escalation to 55 Gy in 25 daily fractions may be accomplished by either photon or proton therapy, with acceptable target coverage and OAR sparing using either modality. However, there is no dosimetric evidence that proton therapy facilitates dose escalation more readily than, as surrounding OARs (e.g., duodenum) receive incrementally higher doses using protons compared to photons. While proton therapy is able to significantly decrease areas receiving low and intermediate dose, the clinical significance of this difference is less certain. This study did not address clinical outcomes of dose escalation by either modality, and the relative importance of high and low doses remains the subject of active investigation. ACKNOWLEDGMENT The authors report no conflicts of interest in conducting the research. a) Author to whom correspondence should be addressed. Electronic mail: Edgar.Ben-Josef@uphs.upenn.edu; Telephone: (215) C. A. Iacobuzio-Donahue et al., DPC4 gene status of the primary carcinoma correlates with patterns of failure in patients with pancreatic cancer, J. Clin. Oncol. 27(11), (29). 2 P. J. Loehrer, Sr. et al., Gemcitabine alone versus gemcitabine plus radiotherapy in patients with locally advanced pancreatic cancer: An Eastern Cooperative Oncology Group trial, J. Clin. Oncol. 29(31), (211). 3 H. M. Ceha et al., Feasibility and efficacy of high dose conformal radiotherapy for patients with locally advanced pancreatic carcinoma, Cancer 89(11), (2). 4 E. Ben-Josef et al., A phase I/II trial of intensity modulated radiation () dose escalation with concurrent fixed-dose rate gemcitabine (FDR- G) in patients with unresectable pancreatic cancer, Int. J. Radiat. Oncol., Biol., Phys. 84(5), (212). 5 J. D. Murphy et al., Full-dose gemcitabine and concurrent radiotherapy for unresectable pancreatic cancer, Int. J. Radiat. Oncol., Biol., Phys. 68(3), (27). 6 D. S. Chang et al., Beam angle selection for intensity-modulated radiotherapy () treatment of unresectable pancreatic cancer: Are noncoplanar beam angles necessary?, Clin. Translat. Oncol. 15(9), (213). 7 E. Ben-Josef et al., Intensity-modulated radiotherapy () and concurrent capecitabine for pancreatic cancer, Int. J. Radiat. Oncol., Biol., Phys. 59(2), (24). 8 M. F. Moyers and D. W. Miller, Range, range modulation, and field radius requirements for proton therapy of prostate cancer, Technol. Cancer Res. Treat. 2(5), (23). 9 P. C. Park et al., A beam-specific planning target volume (PTV) design for proton therapy to account for setup and range uncertainties, Int. J. Radiat. Oncol., Biol., Phys. 82(2), e329 e336 (212). 1 International Commission on Radiation Units and Measurements, Prescribing, recording, and reporting proton-beam therapy, ICRU Report 78 (ICRU, Oxford, 27). 11 R. F. Thompson, RadOnc: An R package for analysis of dose-volume histogram and three-dimensional structural data, J. Rad. Oncol. Informatics 6(1), (214). 12 A. C. Spalding et al., Potential for dose-escalation and reduction of risk in pancreatic cancer using optimization with lexicographic ordering and geud-based cost functions, Med. Phys. 34(2), (27). 13 K. R. Kozak et al., Dosimetric feasibility of hypofractionated proton radiotherapy for neoadjuvant pancreatic cancer treatment, Int. J. Radiat. Oncol., Biol., Phys. 68(5), (27). 14 M. Bouchard et al., Dose escalation with proton or photon radiation treatment for pancreatic cancer, Radiother. Oncol. 92(2), (29). 15 A. Nakamura et al., Analysis of dosimetric parameters associated with acute gastrointestinal toxicity and upper gastrointestinal bleeding in locally advanced pancreatic cancer patients treated with gemcitabine-based concurrent chemoradiotherapy, Int. J. Radiat. Oncol., Biol., Phys. 84(2), (212). 16 P. Kelly et al., Duodenal toxicity after fractionated chemoradiation for unresectable pancreatic cancer, Int. J. Radiat. Oncol., Biol., Phys. 85(3), e143 e149 (213). Medical Physics, Vol. 41, No. 8, August 214

10 Thompson et al.: Proton dosimetry in pancreatic cancer L. A. Dawson et al., The reproducibility of organ position using active breathing control (ABC) during liver radiotherapy, Int. J. Radiat. Oncol., Biol., Phys. 51(5), (21). 18 S. J. Dowdell et al., Monte Carlo study of the potential reduction in out-offield dose using a patient-specific aperture in pencil beam scanning proton therapy, Phys. Med. Biol. 57(1), (212). 19 J. Huang et al., Dose-volume analysis of predictors for gastrointestinal toxicity after concurrent full-dose gemcitabine and radiotherapy for locally advanced pancreatic adenocarcinoma, Int. J. Radiat. Oncol., Biol., Phys. 83(4), (212). 2 P. Prior et al., Consolidating duodenal and small bowel toxicity data via isoeffective dose calculations based on compiled clinical data, Pract. Radiat. Oncol. 4(2), e125 e131 (214). 21 S. Yovino et al., Intensity-modulated radiation therapy significantly improves acute gastrointestinal toxicity in pancreatic and ampullary cancers, Int. J. Radiat. Oncol., Biol., Phys. 79(1), (211). 22 R. C. Nichols, Jr. et al., Proton therapy with concomitant capecitabine for pancreatic and ampullary cancers is associated with a low incidence of gastrointestinal toxicity, Acta Oncol. 52(3), (213). Medical Physics, Vol. 41, No. 8, August 214