Search

Purpose: Deep learning is an emerging technique that allows us to capture imaging information beyond the visually recognizable level of a human being. Because of the anatomic characteristics and location, on-board target verification for radiation delivery to pancreatic tumors is a challenging task. Our goal was to use a deep neural network to localize the pancreatic tumor target on kV x-ray images acquired using an on-board imager for image guided radiation therapy.Methods and Materials: The network is set up in such a way that the input is either a digitally reconstructed radiograph image or a monoscopic x-ray projection image acquired by the on-board imager from a given direction, and the output is the location of the planning target volume in the projection image. To produce a sufficient number of training x-ray images reflecting the vast number of possible clinical scenarios of anatomy distribution, a series of changes were introduced to the planning computed tomography images, including deformation, rotation, and translation, to simulate inter- and intrafractional variations. After model training, the accuracy of the model was evaluated by retrospectively studying patients who underwent pancreatic cancer radiation therapy. Statistical analysis using mean absolute differences (MADs) and Lin's concordance correlation coefficient were used to assess the accuracy of the predicted target positions.Results: MADs between the model-predicted and the actual positions were found to be less than 2.60 mm in anteroposterior, lateral, and oblique directions for both axes in the detector plane. For comparison studies with and without fiducials, MADs are less than 2.49 mm. For all cases, Lin's concordance correlation coefficients between the predicted and actual positions were found to be better than 93%, demonstrating the success of the proposed deep learning for image guided radiation therapy.Conclusions: We demonstrated that markerless pancreatic tumor target localization is achievable with high accuracy by using a deep learning technique approach. [ABSTRACT FROM AUTHOR]

Purpose: To identify prognostic biomarkers in pancreatic cancer using high-throughput quantitative image analysis.Methods and Materials: In this institutional review board-approved study, we retrospectively analyzed images and outcomes for 139 locally advanced pancreatic cancer patients treated with stereotactic body radiation therapy (SBRT). The overall population was split into a training cohort (n=90) and a validation cohort (n=49) according to the time of treatment. We extracted quantitative imaging characteristics from pre-SBRT (18)F-fluorodeoxyglucose positron emission tomography, including statistical, morphologic, and texture features. A Cox proportional hazard regression model was built to predict overall survival (OS) in the training cohort using 162 robust image features. To avoid over-fitting, we applied the elastic net to obtain a sparse set of image features, whose linear combination constitutes a prognostic imaging signature. Univariate and multivariate Cox regression analyses were used to evaluate the association with OS, and concordance index (CI) was used to evaluate the survival prediction accuracy.Results: The prognostic imaging signature included 7 features characterizing different tumor phenotypes, including shape, intensity, and texture. On the validation cohort, univariate analysis showed that this prognostic signature was significantly associated with OS (P=.002, hazard ratio 2.74), which improved upon conventional imaging predictors including tumor volume, maximum standardized uptake value, and total legion glycolysis (P=.018-.028, hazard ratio 1.51-1.57). On multivariate analysis, the proposed signature was the only significant prognostic index (P=.037, hazard ratio 3.72) when adjusted for conventional imaging and clinical factors (P=.123-.870, hazard ratio 0.53-1.30). In terms of CI, the proposed signature scored 0.66 and was significantly better than competing prognostic indices (CI 0.48-0.64, Wilcoxon rank sum test P<1e-6).Conclusion: Quantitative analysis identified novel (18)F-fluorodeoxyglucose positron emission tomography image features that showed improved prognostic value over conventional imaging metrics. If validated in large, prospective cohorts, the new prognostic signature might be used to identify patients for individualized risk-adaptive therapy. [ABSTRACT FROM AUTHOR]

Purpose: To evaluate the feasibility of delivering experimental radiation therapy to tumors in the mouse pancreas. Imaging and treatment were performed using combined CT (computed tomography)/orthovoltage treatment with a rotating gantry.Methods and Materials: After intraperitoneal administration of radiopaque iodinated contrast, abdominal organ delineation was performed by x-ray CT. With this technique we delineated the pancreas and both orthotopic xenografts and genetically engineered disease. Computed tomographic imaging was validated by comparison with magnetic resonance imaging. Therapeutic radiation was delivered via a 1-cm diameter field. Selective x-ray radiation therapy of the noninvasively defined orthotopic mass was confirmed using γH2AX staining. Mice could tolerate a dose of 15 Gy when the field was centered on the pancreas tail, and treatment was delivered as a continuous 360° arc. This strategy was then used for radiation therapy planning for selective delivery of therapeutic x-ray radiation therapy to orthotopic tumors.Results: Tumor growth delay after 15 Gy was monitored, using CT and ultrasound to determine the tumor volume at various times after treatment. Our strategy enables the use of clinical radiation oncology approaches to treat experimental tumors in the pancreas of small animals for the first time. We demonstrate that delivery of 15 Gy from a rotating gantry minimizes background healthy tissue damage and significantly retards tumor growth.Conclusions: This advance permits evaluation of radiation planning and dosing parameters. Accurate noninvasive longitudinal imaging and monitoring of tumor progression and therapeutic response in preclinical models is now possible and can be expected to more effectively evaluate pancreatic cancer disease and therapeutic response. [ABSTRACT FROM AUTHOR]

Purpose The midventilation (midV) approach can be used to take respiratory-induced pancreatic tumor motion into account during radiation therapy. In this study, the dosimetric consequences for organs at risk and tumor coverage of using a midV approach compared with using an internal target volume (ITV) were investigated. Methods and Materials For each of the 18 patients, 2 treatment plans (25 × 2.0 Gy) were created, 1 using an ITV and 1 using a midV approach. The midV dose distribution was blurred using the respiratory-induced motion from 4-dimensional computed tomography. The resulting planning target volume (PTV) coverage for this blurred dose distribution was analyzed; PTV coverage was required to be at least V 95% >98%. In addition, the change in PTV size and the changes in V 10Gy , V 20Gy , V 30Gy , V 40Gy , D mean and D 2cc for the stomach and for the duodenum were analyzed; differences were tested for significance using the Wilcoxon signed-rank test. Results Using a midV approach resulted in sufficient target coverage. A highly significant PTV size reduction of 13.9% ( P <.001) was observed. Also, all dose parameters for the stomach and duodenum, except the D 2cc of the duodenum, improved significantly ( P ≤.002). Conclusions By using the midV approach to account for respiratory-induced tumor motion, a significant PTV reduction and significant dose reductions to the stomach and to the duodenum can be achieved when irradiating pancreatic tumors. [ABSTRACT FROM AUTHOR]

Purpose To determine the optimum sampling strategy for retrospective reconstruction of 4-dimensional (4D) MR data for nonrigid motion characterization of tumor and organs at risk for radiation therapy purposes. Methods and Materials For optimization, we compared 2 surrogate signals (external respiratory bellows and internal MRI navigators) and 2 MR sampling strategies (Cartesian and radial) in terms of image quality and robustness. Using the optimized protocol, 6 pancreatic cancer patients were scanned to calculate the 4D motion. Region of interest analysis was performed to characterize the respiratory-induced motion of the tumor and organs at risk simultaneously. Results The MRI navigator was found to be a more reliable surrogate for pancreatic motion than the respiratory bellows signal. Radial sampling is most benign for undersampling artifacts and intraview motion. Motion characterization revealed interorgan and interpatient variation, as well as heterogeneity within the tumor. Conclusions A robust 4D-MRI method, based on clinically available protocols, is presented and successfully applied to characterize the abdominal motion in a small number of pancreatic cancer patients. [ABSTRACT FROM AUTHOR]

Purpose We report updated outcomes of single- versus multifraction stereotactic body radiation therapy (SBRT) for unresectable pancreatic adenocarcinoma. Methods and Materials We included 167 patients with unresectable pancreatic adenocarcinoma treated at our institution from 2002 to 2013, with 1-fraction (45.5% of patient) or 5-fraction (54.5% of patients) SBRT. The majority of patients (87.5%) received chemotherapy. Results Median follow-up was 7.9 months (range: 0.1-63.6). The 6- and 12-month cumulative incidence rates (CIR) of local recurrence for patients treated with single-fraction SBRT were 5.3% (95% confidence interval [CI], 0.2%-10.4%) and 9.5% (95% CI, 2.7%-16.2%), respectively. The 6- and 12-month CIR with multifraction SBRT were 3.4% (95% CI, 0.0-7.2%) and 11.7% (95% CI, 4.8%-18.6%), respectively. Median survival from diagnosis for all patients was 13.6 months (95% CI, 12.2-15.0 months). The 6- and 12- month survival rates from SBRT for the single-fraction group were 67.0% (95% CI, 57.2%-78.5%) and 30.8% (95% CI, 21.9%-43.6%), respectively. The 6- and 12- month survival rates for the multifraction group were 75.7% (95% CI, 67.2%-85.3%) and 34.9% (95% CI, 26.1%-46.8%), respectively. There were no differences in CIR or survival rates between the single- and multifraction groups. The 6- and 12-month cumulative incidence rates of gastrointestinal toxicity grade ≥3 were 8.1% (95% CI, 1.8%-14.4%) and 12.3% (95% CI, 4.7%-20.0%), respectively, in the single-fraction group, and both were 5.6% (95% CI, 0.8%-10.5%) in the multifraction group. There were significantly fewer instances of toxicity grade ≥2 with multifraction SBRT ( P =.005). Local recurrence and toxicity grade ≥2 were independent predictors of worse survival. Conclusions Multifraction SBRT for pancreatic cancer significantly reduces gastrointestinal toxicity without compromising local control. [ABSTRACT FROM AUTHOR]

Purpose: Because of low soft-tissue contrast of cone beam computed tomography (CBCT), fiducial markers are often used for radiation therapy patient setup verification. For pancreatic cancer patients, biliary stents have been suggested as surrogate fiducials. Using intratumoral fiducials as standard for tumor position, this study aims to quantify the suitability of biliary stents for measuring interfractional and respiratory-induced position variations of pancreatic tumors. Methods and Materials: Eleven pancreatic cancer patients with intratumoral fiducials and a biliary stent were included in this study. Daily CBCT scans (243 in total) were registered with a reference CT scan, based on bony anatomy, on fiducial markers, and on the biliary stent, respectively. We analyzed the differences in tumor position (ie, markers center-of-mass position) among these 3 registrations. In addition, we measured for 9 patients the magnitude of respiratory-induced motion (MM) of the markers and of the stent on 4-dimensional CT (4DCT) and determined the difference between these 2 magnitudes (ΔMM). Results: The stent indicated tumor position better than bony anatomy in 67% of fractions; the absolute difference between the markers and stent registration was >5 mm in 46% of fractions and >10 mm in 20% of fractions. Large PTV margins (superior-inferior direction, >19 mm) would be needed to account for this interfractional position variability. On 4DCT, we found in superior-inferior direction a mean ΔMM of 0.5 mm (range, –2.6 to 4.2 mm). Conclusions: For respiratory-induced motion, the mean ΔMM is small, but for individual patients the absolute difference can be >4 mm. For interfractional position variations, a stent is, on average, a better surrogate fiducial than bony anatomy, but large PTV margins would still be required. Therefore, intratumoral fiducials are recommended for online setup verification for all pancreatic patients scheduled for radiation therapy, including patients with a biliary stent. [Copyright &y& Elsevier]

Purpose: We assessed the accuracy of abdominal magnetic resonance imaging (MRI) for determining tumor size by comparing the preoperative contrast-enhanced T1-weighted gradient echo (3-dimensional [3D] volumetric interpolated breath-hold [VIBE]) MRI tumor size with pathologic specimen size. Methods and Materials: The records of 92 patients who had both preoperative contrast-enhanced 3D VIBE MRI images and detailed pathologic specimen measurements were available for review. Primary tumor size from the MRI was independently measured by a single diagnostic radiologist (P.M.) who was blinded to the pathology reports. Pathologic tumor measurements from gross specimens were obtained from the pathology reports. The maximum dimensions of tumor measured in any plane on the MRI and the gross specimen were compared. The median difference between the pathology sample and the MRI measurements was calculated. A paired t test was conducted to test for differences between the MRI and pathology measurements. The Pearson correlation coefficient was used to measure the association of disparity between the MRI and pathology sizes with the pathology size. Disparities relative to pathology size were also examined and tested for significance using a 1-sample t test. Results: The median patient age was 64.5 years. The primary site was pancreatic head in 81 patients, body in 4, and tail in 7. Three patients were American Joint Commission on Cancer stage IA, 7 stage IB, 21 stage IIA, 58 stage IIB, and 3 stage III. The 3D VIBE MRI underestimated tumor size by a median difference of 4 mm (range, −34-22 mm). The median largest tumor dimensions on MRI and pathology specimen were 2.65 cm (range, 1.5-9.5 cm) and 3.2 cm (range, 1.3-10 cm), respectively. Conclusions: Contrast-enhanced 3D VIBE MRI underestimates tumor size by 4 mm when compared with pathologic specimen. Advanced abdominal MRI sequences warrant further investigation for radiation therapy planning in pancreatic adenocarcinoma before routine integration into the treatment planning process. [Copyright &y& Elsevier]