Dalong Pang

Fully Automated Simultaneous Integrated Boosted–Intensity Modulated Radiation Therapy Treatment Planning Is Feasible for Head-and-Neck Cancer: A Prospective Clinical Study

PURPOSE To prospectively determine whether overlap volume histogram (OVH)-driven, automated simultaneous integrated boosted (SIB)-intensity-modulated radiation therapy (IMRT) treatment planning for head-and-neck cancer can be implemented in clinics. METHODS AND MATERIALS A prospective study was designed to compare fully automated plans (APs) created by an OVH-driven, automated planning application with clinical plans (CPs) created by dosimetrists in a 3-dose-level (70 Gy, 63 Gy, and 58.1 Gy), head-and-neck SIB-IMRT planning. Because primary organ sparing (cord, brain, brainstem, mandible, and optic nerve/chiasm) always received the highest priority in clinical planning, the study aimed to show the noninferiority of APs with respect to PTV coverage and secondary organ sparing (parotid, brachial plexus, esophagus, larynx, inner ear, and oral mucosa). The sample size was determined a priori by a superiority hypothesis test that had 85% power to detect a 4% dose decrease in secondary organ sparing with a 2-sided alpha level of 0.05. A generalized estimating equation (GEE) regression model was used for statistical comparison. RESULTS Forty consecutive patients were accrued from July to December 2010. GEE analysis indicated that in APs, overall average dose to the secondary organs was reduced by 1.16 (95% CI = 0.09-2.33) with P=.04, overall average PTV coverage was increased by 0.26% (95% CI = 0.06-0.47) with P=.02 and overall average dose to the primary organs was reduced by 1.14 Gy (95% CI = 0.45-1.8) with P=.004. A physician determined that all APs could be delivered to patients, and APs were clinically superior in 27 of 40 cases. CONCLUSIONS The application can be implemented in clinics as a fast, reliable, and consistent way of generating plans that need only minor adjustments to meet specific clinical needs.

Using overlap volume histogram and IMRT plan data to guide and automate VMAT planning: A head‐and‐neck case study

PURPOSE To investigate whether an overlap volume histogram (OVH)-driven planning application using an intensity-modulated radiation therapy (IMRT) database can guide and automate volumetric-modulated arc therapy (VMAT) planning for head-and-neck cancer. METHODS Based on comparable head-and-neck dosimetric results between planner-generated VMAT and IMRT plans, an inhouse developed, OVH-driven automated planning application containing a database of prior clinical head-and-neck IMRT plans is built into Pinnacle(3) SmartArc for VMAT planning. Double-arc VMAT plans of four oropharynx, four nasopharynx, and four larynx patients are generated and compared with corresponding clinical IMRT plans. RESULTS Each VMAT plan is automatically generated in two optimization rounds, while the average number of optimization rounds in generating a clinical IMRT plan is 43. In VMAT plans, statistical superiority (p < 0.01) in sparing of the cord+4 mm, brainstem, brachial plexus, larynx, and inner ear is observed with a slight degradation in low-dose-level planning target volume (PTV) coverage. On average, D(0.1 cc) to the cord+4 mm, brainstem and brachial plexus is reduced by 3.7, 4.9, and 1.6 Gy, respectively; V(50 Gy) to the larynx is reduced by 5.3%; mean dose to the inner ear is reduced by 4.4 Gy; V(95) of low-dose-level PTV coverage is reduced by 0.3% with p = 0.25. CONCLUSIONS IMRT-data-driven VMAT planning offers a potential method for generating VMAT plans that are comparable to IMRT plans in terms of dosimetric quality.

Investigation of neutron-induced damage in DNA by atomic force microscopy: experimental evidence of clustered DNA lesions.

Using atomic force microscopy (AFM), we have investigated neutron-induced DNA double-strand breaks in plasmids in aqueous solution. AFM permits direct measurement of individual DNA molecules with an accuracy of a few nanometers. Furthermore, the analysis of the DNA fragment size distribution is non-parametric, whereas other methods are dependent on the model. Neutron irradiation of DNA results in the generation of many short fragments, an observation not made for damage induced by low-LET radiation. These data provide clear experimental evidence for the existence of clustered DNA double-strand breaks and demonstrate that short DNA fragments may be produced by such radiations in the absence of a nucleosomal DNA structure.

Spatial Distribution of Radiation-Induced Double-Strand Breaks in Plasmid DNA as Resolved by Atomic Force Microscopy

Abstract Pang, D., Rodgers, J. E., Berman, B. L., Chasovskikh, S. and Dritschilo, A. Spatial Distribution of Radiation-Induced Double-Strand Breaks in Plasmid DNA as Resolved by Atomic Force Microscopy. Radiat. Res. 164, 755–765 (2005). Atomic force microscopy (AFM) has been used to directly visualize, size and compare the DNA fragments resulting from exposure to low- and high-LET radiation. Double-stranded pUC-19 plasmid (“naked”) DNA samples were irradiated by electron-beam or reactor neutron fluxes with doses ranging from 0.9 to 10 kGy. AFM scanning in the tapping mode was used to image and measure the DNA fragment lengths (ranging from a few bp up to 2864 bp long). Double-strand break (DSB) distributions resulting from high-LET neutron and lower-LET electron irradiation revealed a distinct difference between the effects of these two types of radiation: Low-LET radiation-induced DSBs are distributed more uniformly along the DNA, whereas a much larger proportion of neutron-induced DSBs are distributed locally and densely. Furthermore, comparisons with predictions of a random DSB model of radiation damage show that neutron-induced DSBs deviate more from the model than do electron-induced DSBs. In summary, our high-resolution AFM measurements of radiation-induced DNA fragment-length distributions reveal an increased number of very short fragments and hence clustering of DSBs induced by the high-LET neutron radiation compared with low-LET electron radiation and a random DSB model prediction.

Improved robotic stereotactic body radiation therapy plan quality and planning efficacy for organ-confined prostate cancer utilizing overlap-volume histogram-driven planning methodology.

BACKGROUND AND PURPOSE This study is to determine if the overlap-volume histogram (OVH)-driven planning methodology can be adapted to robotic SBRT (CyberKnife Robotic Radiosurgery System) to further minimize the bladder and rectal doses achieved in plans manually-created by clinical planners. METHODS AND MATERIALS A database containing clinically-delivered, robotic SBRT plans (7.25 Gy/fraction in 36.25 Gy) of 425 patients with localized prostate cancer was used as a cohort to establish an organs distance-to-dose model. The OVH-driven planning methodology was refined by adding the PTV volume factor to counter the targets dose fall-off effect and incorporated into Multiplan to automate SBRT planning. For validation, automated plans (APs) for 12 new patients were generated, and their achieved dose/volume values were compared to the corresponding manually-created, clinically-delivered plans (CPs). A two-sided, Wilcoxon rank-sum test was used for statistical comparison with a significance level of p<0.05. RESULTS PTVs V(36.25 Gy) was comparable: 95.6% in CPs comparing to 95.1% in APs (p=0.2). On average, the refined approach lowered V(18.12 Gy) to the bladder and rectum by 8.2% (p<0.05) and 6.4% (p=0.14). A physician confirmed APs were clinically acceptable. CONCLUSIONS The improvements in APs could further reduce toxicities observed in SBRT for organ-confined prostate cancer.

Significant disparity in base and sugar damage in DNA resulting from neutron and electron irradiation

In this study, a comparison of the effects of neutron and electron irradiation of aqueous DNA solutions was investigated to characterize potential neutron signatures in DNA damage induction. Ionizing radiation generates numerous lesions in DNA, including base and sugar lesions, lesions involving base–sugar combinations (e.g. 8,5′-cyclopurine-2′-deoxynucleosides) and DNA–protein cross-links, as well as single- and double-strand breaks and clustered damage. The characteristics of damage depend on the linear energy transfer (LET) of the incident radiation. Here we investigated DNA damage using aqueous DNA solutions in 10 mmol/l phosphate buffer from 0–80 Gy by low-LET electrons (10 Gy/min) and the specific high-LET (∼0.16 Gy/h) neutrons formed by spontaneous 252Cf decay fissions. 8-hydroxy-2′-deoxyguanosine (8-OH-dG), (5′R)-8,5′-cyclo-2′-deoxyadenosine (R-cdA) and (5′S)-8,5′-cyclo-2′-deoxyadenosine (S-cdA) were quantified using liquid chromatography–isotope-dilution tandem mass spectrometry to demonstrate a linear dose dependence for induction of 8-OH-dG by both types of radiation, although neutron irradiation was ∼50% less effective at a given dose compared with electron irradiation. Electron irradiation resulted in an exponential increase in S-cdA and R-cdA with dose, whereas neutron irradiation induced substantially less damage and the amount of damage increased only gradually with dose. Addition of 30 mmol/l 2-amino-2-(hydroxymethyl)-1,3-propanediol (TRIS), a free radical scavenger, to the DNA solution before irradiation reduced lesion induction to background levels for both types of radiation. These results provide insight into the mechanisms of DNA damage by high-LET 252Cf decay neutrons and low-LET electrons, leading to enhanced understanding of the potential biological effects of these types of irradiation.

SU‐E‐T‐67: A Simple Tool for Evaluation of Fiducial Placement for CyberKnife SRS/SBRT Treatment

Purpose: To develop a simple software tool to help determine whether a fiducial placement meets the Spacing and Collinearity criteria for CyberKnife 6D target tracking. Methods: An Excel spreadsheet has been developed for this purpose. The spreadsheet calculates distances and angles for all possible combinations of fiducial triplets based on the fiducial coordinates entered. In general, a group of 4 fiducials forms 4 triplets; a group of 6 fiducials forms 20 triplets. Distance between any two points in 3D space can be calculated easily. Angles formed by lines joining a fiducial triplet are calculated by using the law of cosines. Calculations using scalar product of vectors method are also implemented and used as a redundancy check. Fiducial placement and CyberKnife treatment data from 78 prostate patients were analyzed. Results: The spreadsheet checks each fiducial triplet against the Spacing & Collinearity Thresholds. Distances or angles failed to meet the thresholds are flagged. If none of the fiducial triplets meet the criteria, extra fiducials need to be implanted. Of the 78 patients studied, most patients were implanted with 4 fiducials; seven patients had 3, and twenty patients had 6 fiducials. A total of the 345 fiducials and 600+ triplets were analyzed. The study showed that 72% of the fiducial distances exceeded the 20‐mm Spacing Threshold, and 99% of the fiducial angles exceeded the 15‐degree Collinearity Threshold. Conclusions: We have developed an Excel spreadsheet to verify that a fiducial placement meets the Cyberknife 6D tracking criteria before patient treatment.Analyzing past patient data helps provide guidance on how fiducial placement may be improved. It is important that least one of the fiducial triplets exceed both Spacing and Collinearity Thresholds, as the study has found that there were cases where a fiducial triplet met the Spacing criteria but failed Collinearity criteria, and vice versa.

SU‐E‐T‐670: Using Overlap Volume Histogram Analysis of a Prior Plan Dataset to Generate Clinically Acceptable Plans for CyberKnife Robotic Radiosurgery Treatment of Localized Prostate Cancer

PURPOSE CyberKnife offers the potential benefits of non-isocentric, non-coplanar treatment delivery. However, its planning is a laborious manual, trial-and-error process. This study is to investigate whether an overlap volume histogram (OVH)-driven planning approach can produce clinically acceptable plans for treatment of localized prostate cancer. METHODS It is assumed that given consistent target coverage, patients with a relevant target-organ spatial relationship should have similar organ sparing. The OVH is used to characterize the 3-D spatial relationship between an organ and a target. A database containing the OVH and DVH of prior plans is built to serve as an external reference. During the initial planning, the OVH is used to search through the database to find a prior patient group whose target-organ relationship is related to that of a new patient. The planning objectives for the new patient are then estimated from the group and input into the CyberKnife TPS for optimization. To demonstrate the effectiveness of the method, the plans of 12 prostate patients (prescription: 36.25Gy in 5 fractions) are generated by the OVH approach and compared to the corresponding clinical plans using the in-house dosimetric guidelines. RESULTS Physicians confirm that OVH plans are clinically acceptable. OVH plans: on average, Vcc(37Gy) and V(18.12Gy) to the bladder decrease 1cc and 6% (p<0.05); Vcc(36Gy) and V(18.12Gy) to the rectum decrease 0.02cc and 4.3% (p=0.3); Vcc(40Gy) to the prostatic urethra decreases 0.08cc (p=0.01); V(14.5Gy) to the femur heads decreases 0.58% (p=0.01). V(37Gy) to the membranous urethra increases from 23.1% to 24.5% (p=0.75); V(29.5Gy) to the penile bulb increases from 5.5% to 10.7% (p=0.35); V(36.25Gy) to PTV increases 0.2% (p=0.4); the estimated delivery time deceases 3.5 minutes (p=0.01). CONCLUSION With respect to planner-created clinical plans, this approach offers an alternative way to generate clinically acceptable plans. It advances the possibility of automated CyberKnife planning. C. Sims, employed by Accuray, Inc;S.P Collins, Accuray clinical consultant.

SU‐FF‐T‐111: Head‐And‐Neck IMRT Without Beam‐Splitting

Purpose: Varian MLC leaf travel is limited to 14.5 cm. For large PTV, an IMRT field will be split into two or three sub‐fields. Thus a 9‐beam head‐and‐neck plan may end up with 18 or more treatment fields. The purpose of this study is to develop a non‐split IMRT planning technique and compare the quality of non‐split plans with beam‐split counterparts. Method: Varian Eclipse user can choose fixed‐jaws and set field‐size less than 14.5 cm so that IMRT field will not split. A small part of the PTV may be blocked by x‐jaws in a particular beam, but the missing dose from one beam can be covered by other beams at different gantry angles. We compared our previously treated 9‐beam‐split‐18‐field head‐and‐neck IMRT plans with corresponding non‐split plans of 9, 11, and 15 gantry angles. Plan DVH and IMRT QA results were analyzed. Results: Non‐split plans produced the same dose coverage as beam‐split plans. DVH curves of PTV for split and non‐split plans almost overlapped with each other. DVH of organs‐at‐risk were slightly different. Total treatment MU were about the same for all the plans. All IMRT QA plans delivered on a Varian Trilogy passed physics QA criteria, with non‐split plans showing slightly better passing scores. Conclusions: Creating IMRT plans without beam‐splitting is encouraged. Non‐split IMRT plans can be as good as beam‐split plans, but use only half of the number of beams. Quality of non‐split plans improves with increasing number of beams. The amount of improvement was larger for the number of beams going from 9 to 11 than that from 13 to 15, demonstrating a balance of costs and benefits. Cutting the number of beams by half may result in a number of benefits: more accurate dose delivery, shorter treatment time, and increased machine throughput and productivity.