Search

BACKGROUND The addition of chemotherapy to adjuvant radiotherapy (chemotherapy and radiation therapy [CRT]) improves overall survival (OS) for patients with high-risk grade 2 gliomas; however, the impact of chemotherapy alone (CA) is unknown. This study compares the OS of patients with high-risk grade 2 gliomas treated with CA versus CRT. METHODS Patients with high-risk grade 2 gliomas (subtotal resection or age ≥ 40 years) with oligodendrogliomas, astrocytomas, or mixed tumors were identified with the National Cancer Data Base. Patients were grouped into CA and CRT cohorts. Univariate analyses and multivariate analyses (MVAs) were performed. Propensity score (PS) matching was also implemented. The Kaplan-Meier method was used to analyze OS. RESULTS A total of 1054 patients with high-risk grade 2 gliomas were identified: 496 (47.1%) received CA, and 558 (52.9%) received CRT. Patients treated with CA were more likely (all P values 6 cm, astrocytoma histology, and older age were predictors for worse OS (all P values

Treatment planning for whole‐brain radiation treatment is technically a simple process, but in practice it takes valuable clinical time of repetitive and tedious tasks. This report presents a method that automatically segments the relevant target and normal tissues, and creates a treatment plan in only a few minutes after patient simulation. Segmentation of target and critical structures is performed automatically through morphological operations on the soft tissue and was validated by comparing with manual clinical segmentation using the Dice coefficient and Hausdorff distance. The treatment plan is generated by searching a database of previous cases for patients with similar anatomy. In this search, each database case is ranked in terms of similarity using a customized metric designed for sensitivity by including only geometrical changes that affect the dose distribution. The database case with the best match is automatically modified to replace relevant patient info and isocenter position while maintaining original beam and MLC settings. Fifteen patients with marginally acceptable treatment plans were used to validate the method. In each of these cases the anatomy was accurately segmented, but the beams and MLC settings led to a suboptimal treatment plan by either underdosing the brain or excessively irradiating critical normal tissues. For each case, the anatomy was automatically segmented with the proposed method, and the automated and manual segmentations were then compared. The mean Dice coefficient was 0.97, with a standard deviation of 0.008 for the brain, 0.85±0.009 for the eyes, and 0.67±0.11 for the lens. The mean Euclidian distance was 0.13±0.13 mm for the brain, 0.27±0.31 for the eye, and 2.34±7.23 for the lens. Each case was then subsequently matched against a database of 70 validated treatment plans and the best matching plan (termed autoplanned), was compared retrospectively with the clinical plans in terms of brain coverage and maximum doses to critical structures. Maximum doses were reduced by a maximum of 8.37 Gy for the left eye (mean 2.08), 11.67 for the right eye (1.90) and, respectively, 25.44 (5.59) for the left lens and 24.40 (4.85) for the right lens. Time to generate the autoplan, including the segmentation, was 3−4 min. Automated database‐ based matching is an alternative to classical treatment planning that improves quality while providing a cost‐effective solution to planning through modifying previous validated plans to match a current patient's anatomy. PACS number: 87.55.D, 87.55.tg, 87.57.nm

BACKGROUND The purpose of this study was to evaluate predictors of early distant brain failure (DBF) and salvage whole-brain radiotherapy (WBRT) after treatment with stereotactic radiosurgery (SRS) for brain metastases and create a clinically relevant risk score to stratify patients’ risk for these events. METHODS The records of 270 patients with brain metastases who were treated with SRS between 2003 and 2012 were reviewed. Pretreatment patient and tumor characteristics were analyzed with univariate and multivariate analyses. The cumulative incidences of first DBF and salvage WBRT were calculated. Significant factors were used to create a score for stratifying early (6-month) DBF risk. RESULTS No prior WBRT, a total lesion volume

BACKGROUND Bevacizumab is a monoclonal antibody targeting vascular endothelial growth factor and is approved for the treatment of patients with recurrent glioblastoma (GBM). Previous authors have reported differential response to bevacizumab on an individual basis. Bevacizumab-induced hypertension is a well-documented side effect, and some reports have suggested this occurrence to be related to treatment outcome in other cancers. In the current study, the authors analyzed patients with recurrent GBM who were treated with bevacizumab based on whether the patients developed drug-induced hypertension. METHODS All patients with GBM treated within the Emory Healthcare system from 2007 through 2012 were reviewed. A total of 82 patients were identified who received bevacizumab for the treatment of recurrent GBM and were included in the current study. Patients were classified as normotensive or hypertensive depending on whether hypertension developed that was attributable to therapy. Progression-free survival (PFS) and overall survival (OS) were graphed by the Kaplan-Meier method. Univariate and multivariate analyses were performed using the Cox proportional hazards method. RESULTS The median follow-up was 19.7 months. Of the 82 patients with recurrent GBM who were treated with bevacizumab, 30 developed drug-induced hypertension. The median time to the development of hypertension was 21 days. The median PFS for the normotensive and hypertensive groups were 2.5 months (95% confidence interval [95% CI], 1.6-3.0 months) and 6.7 months (95% CI, 4.6-10.0 months), respectively (P

BACKGROUND The patterns of lobar involvement, optimal treatment, and disease course among patients with gliomatosis cerebri (GC) have not been fully characterized. The current study evaluates the clinical presentations and outcomes for patients with GC treated with radiation therapy (RT) at our institution. METHODS A total of 26 patients (25 with follow-up) with GC were diagnosed and treated between January 2004 and June 2012. Inclusion criteria consisted of brain magnetic resonance imaging and neuroradiology confirmation of contiguous involvement of ≥ 3 lobes/lobar equivalents with preservation of neural architecture. Patients were treated with either partial-brain RT to involved tumor (25 patients) or whole-brain RT (1 patient). The median RT dose was 54.0 Gray. The median follow-up was 17.3 months. RESULTS The median age of the patients at the time of diagnosis was 57 years. Twenty-one patients (81%) and 5 patients (19%) had 3 to 6 and ≥ 7 involved lobes/lobar equivalents, respectively. The median progression-free survival and overall survival were 7.4 months and 14.9 months, respectively. Fifteen patients experienced radiographic disease progression after partial-brain RT, 14 of whom (93%) developed infield disease recurrence. On univariate analysis, higher tumor grade and type II GC (with focal mass) were associated with a poorer progression-free survival. The extent of lobar involvement and chemotherapy were not associated with overall survival. CONCLUSIONS Even with partial-brain RT, nearly all disease recurrences were infield and clinical outcomes were similar to previous GC series, thereby suggesting that whole-brain RT is not necessary for this patient population. A greater number of involved lobes did not correlate with inferior outcomes. Further studies are necessary to establish more uniform and optimal treatments for this rare disease. Cancer 2014;120:2713–2720. © 2014 American Cancer Society.

Frameless radiosurgery is an attractive alternative to the framed procedure if it can be performed with comparable precision in a reasonable time frame. Here, we present a positioning approach for frameless radiosurgery based on in‐room volumetric imaging coupled with an advanced six‐degrees‐of‐freedom (6 DOF) image registration technique which avoids use of a bite block. Patient motion is restricted with a custom thermoplastic mask. Accurate positioning is achieved by registering a cone‐beam CT to the planning CT scan and applying all translational and rotational shifts using a custom couch mount. System accuracy was initially verified on an anthropomorphic phantom. Isocenters of delineated targets in the phantom were computed and aligned by our system with an average accuracy of 0.2 mm, 0.3 mm, and 0.4 mm in the lateral, vertical, and longitudinal directions, respectively. The accuracy in the rotational directions was 0.1°, 0.2°, and 0.1° in the pitch, roll, and yaw, respectively. An additional test was performed using the phantom in which known shifts were introduced. Misalignments up to 10 mm and 3° in all directions/rotations were introduced in our phantom and recovered to an ideal alignment within 0.2 mm, 0.3 mm, and 0.4 mm in the lateral, vertical, and longitudinal directions, respectively, and within 0.3° in any rotational axis. These values are less than couch motion precision. Our first 28 patients with 38 targets treated over 63 fractions are analyzed in the patient positioning phase of the study. Mean error in the shifts predicted by the system were less than 0.5 mm in any translational direction and less than 0.3° in any rotation, as assessed by a confirmation CBCT scan. We conclude that accurate and efficient frameless radiosurgery positioning is achievable without the need for a bite block by using our 6 DOF registration method. This system is inexpensive compared to a couch‐based 6 DOF system, improves patient comfort compared to systems that utilize a bite block, and is ideal for the treatment of pediatric patients with or without general anesthesia, as well as of patients with dental issues. From this study, it is clear that only adjusting for 4 DOF may, in some cases, lead to significant compromise in PTV coverage. Since performing the additional match with 6 DOF in our registration system only adds a relatively short amount of time to the overall process, we advocate making the precise match in all cases. PACS number: 87.55.tm; 87.55.Qr; 87.57.nj

An automated optimization algorithm based on mixed integer programming techniques is presented for generating high-quality treatment plans for LINAC radiosurgery treatment. The physical planning in radiosurgery treatment involves selecting among a large collection of beams with different physical parameters an optimal beam configuration (geometries and intensities) to deliver the clinically prescribed radiation dose to the tumor volume while sparing the nearby critical structure and normal tissue. The proposed mixed integer programming models incorporate strict dose restrictions on tumor volume, and constraints on the desired number of beams, isocenters, couch angles, and gantry angles. The model seeks to deliver full prescription dose coverage and uniform radiation dose to the tumor volume while minimizing the excess radiation to the periphery normal tissue. In particular, it ensures that proximal normal tissues receive minimal dose via rapid dose fall-off. Preliminary numerical tests on a single patient case indicate that this approach can produce exceptionally high-quality plans in a fraction of the time required using the procedure currently employed by clinicians. The resulting plans provide highly uniform prescription dose to the tumor volume while drastically reducing the irradiation received by the proximal critical normal tissue.

Background: Restenosis after balloon angioplasty remains a substantial clinical problem. Endovascular radiation has been shown to reduce neointima formation in animal models of restenosis, and the results of early clinical trials are promising. However, this technique has several limitations, particularly, the inability to delay or fractionate therapy and inhomogeneous dose distribution. These shortcomings might be overcome by delivering radiation therapy from an external approach. This study investigated the effects of high dose external beam irradiation on balloon-injured pig coronary arteries. Methods and Results: Eight domestic juvenile pigs underwent balloon overstretch injury. External beam irradiation (21 Gy) of the heart was performed within 2 hours on four pigs and the remainder were not irradiated (controls). Sections of paraffin embedded injured arteries, perfusion fixed at 4 weeks, were measured by computer-assisted histomorphometry. Luminal area in the 21-Gy group was 3.84 ± 1.80 mm2 compared to 2.0 4 ± 0.49 mm2 in the controls (P < 0.01). Neointimal area and maximal intimal thickness were reduced in the 21-Gy group (0.23 ± 0.36 vs 1.36 ± 0.67mm2, P < 0.0005; 0.25 ± 0.35 vs 0.64 ± 0.2, P < 0.01, respectively). Diffuse focal interstitial fibrosis was observed in the myocardium of the irradiated group. Conclusions: At 21 Gy, external beam irradiation significantly increased lumen size and reduced neointima formation in balloon-injured pig coronary arteries. However, myocardial damage was also induced by irradiation of the whole heart at this dose.

Recent preclinical and clinical studies indicate that irradiation using ionizing radiation in the dose range of 15 to 30 Gy may reduce the occurrence of restenosis in patients who have undergone an angioplasty. Several delivery systems of intravascular brachytherapy have been developed to deliver radiation doses in this range with minimal normal tissue toxicity. In late 1995 the American Association of Physicists in Medicine (AAPM) formed a task group to investigate these issues and to report the current state of the art of intravascular brachytherapy physics. The report of this task group is presented here.

Vascular constriction post PTCA is a major component in the mechanism of restenosis following intervention. Ionizing radiation demonstrated reduction of neointima formation after injury in animal models and lowered the restenosis rates in pilot clinical studies. To determine the effect of intracoronary radiation therapy on vascular remodeling, angiograms from two radiation trials were analyzed by QCA methods. Patients in these trials had de novo lesions and were treated with balloon angioplasty followed by either beta or gamma radiation. All patients were studied angiographically at 6 months; patients with total occlusion at the treated artery were excluded from the analysis. In the gamma trial, 192-Iridium was utilized in 14 patients (15 lesions) with doses between 20-25 Gy. In the beta trial, 90-Sr/Y was utilized in 17 patients (17 lesions) with doses between 12-16 Gy. The QCA analysis from these studies demonstrated negative late loss and late loss index at six months for patients from the beta (-0.02 ± 0.3) and the gamma (-0.19 ± 0. 3) study. The effect of positive remodeling was maintained at 24 months, -0.16 ± 0.4 in the gamma group. Larger MLD at follow-up compared to the immediate post MLD were demonstrated in 50% of the patients from both studies. Thus, intracoronary radiation resulted in lower late loss and late loss index rates than previously reported following balloon angioplasty alone suggesting a positive vascular remodeling effect of intracoronary radiation.

Endocrine dysfunction can be challenging to diagnose in children treated for brain tumors. Treatments are available for hormonal replacement and when necessary, hormonal suppression. Without these endocrine treatment regimens, life can be unnecessarily difficult or unpleasant. An endocrine survey can be used to screen at-risk neuro-oncology patients once or twice a year to facilitate the recognition of endocrine dysfunction. It is hoped that through the use of a routine screening program, physicians will be able to diagnose and begin treatment of endocrine problems in a time-efficient manner.

This study examines the characteristics of the patient groups, treatment parameters, and results of therapy for 126 intracranial lesions treated with stereotactic radiosurgery utilizing a "patient rotator" and a linear accelerator. A retrospective review was conducted and data on 122 patients (21 patients with arteriovenous malformations-AVMs, 40 patients with 41 metastatic tumors, 24 patients with malignant gliomas and 37 patients with other benign lesions) were analyzed. Clinical and radiographic response was obtained from chart review and/or telephone follow-up. The average follow-up was 12.1, 13.0, 5.7, and 23.1 months in patients with AVMs, malignant gliomas, metastases, and other lesions, respectively. Median survival times (MST) of the metastatic and glioma group were 9 and 38 months, respectively. Complete or partial radiographic response at follow-up was seen in 62.5% of AVM patients, 33.3% of patients with metastases, 11.8% of patients with malignant gliomas, and 19.3% of patients with other lesions. (1) Local control of brain metastases and benign intracranial lesions can be obtained with single dose stereotactic radiosurgery. (2) The MST and local control rate for metastatic disease obtained are comparable with those in the literature. (3) The patient rotator method for stereotactic radiosurgery is an effective tool for treating selected intracranial lesions.

Purpose: Treatment plan compliance to physician's intent is still assessed by hand in clinical practice by visually comparing prescription details against corresponding plan values. We report on a script that uses the Varian's programming interface to automate the scoring of treatment plan dose values to the physician's prescribed constraints. Methods: The Eclipse Scripting Advance Programming Interface (ESAPI) is a FDA-approved scripting environment written in C# that allows programming access to a patient's record in Aria and Eclipse. With the ESAPI, it is possible to create customized add-on software libraries that can be installed in the planning system as a pull-down menu item. We used the flexibility of ESAPI to parse plan quality metrics from the prescription and compare them automatically against the corresponding plan values. The script can check fractionation as well as minimum, mean and maximum doses to an organ specified in the prescription, DVHs metrics such as volume or dose coverage for both targets and critical organs, conformity indexes specific to radio surgery and biological indexes such as EUD, TCP and NCTP. Lists of generally accepted constraints limits, such as QUANTEC or clinical trials constraints are implemented in the script and can be invoked to automatically fill the prescription. Results: The script interprets the prescription and compares plan quality indices parsed from it against the counterpart values extracted from the plan. A coloring scheming is employed to automatically mark with green, yellow or red plan quality metrics that meet, are borderline fail (10%) or fail the prescription criteria. While manually performing this task can take 30 minutes for a complex head and neck case with many constraints, the script takes approximately 1 minute. Conclusion: The script provided easy to understand feedback on quality metrics in a simple and practical solution that significantly streamline plan double-checks in clinical practice.

Between May 1986 and February 1988, 57 patients were accrued to the second dose level of a Phase I/II Pediatric Oncology Group (POG) study exploring the use of hyperfractionated radiation therapy (HRT) in children with high-risk brain stem tumors. Local fields were treated with fraction sizes of 117 cGy given twice daily, with a minimum interfraction interval of 6 hours, to a total dose of 7020 cGy in 60 fractions over 6 weeks. Information regarding clinical status during HRT was available for 55 patients (44 [80%] improved, 6 remained stable, and 5 deteriorated). Results of initial and follow-up computed tomography (CT) scan and/or magnetic resonance imaging (MRI) were available for review for 52 patients. One patient had a complete response (CR) to treatment, 3 had a partial response (PR) (more than 50% response), and 40 remained stable, for a total response rate (CR + PR + stable) of 77%. Median time to disease progression was 6 months. Median survival time was 10 months. Survival rate was 39.6% (standard error [SE] = 6.6%) at 1 year and 23% (SE = 5.8%) at 2 years. Complications of treatment included an enhanced skin reaction in six patients and otitis media and/or externa in nine. One patient bled into tumor shortly after completion of HRT, and three had intralesional necrosis. Five patients continued taking steroids for protracted periods in the face of improved clinical and/or radiologic findings. Complications related to the use of steroids included opportunistic infections, impaired glucose tolerance, hypertension, osteoporosis, and significant mood changes. In no patient was there evidence of any late injury attributable to HRT. When compared with results of treatment with HRT at a lower dose level (6600 cGy), there appears to be a trend toward improved survival at 7020 cGy despite a less favorable patient population at the higher dose level. A second dose escalation to 7560 cGy in 60 fractions over 6 weeks has been implemented as planned.

Purpose: We implemented high resolution, 3D whole brain MR spectroscopic imaging (MRSI) technology to provide additional information on glioblastoma (GBM) brain tumors not readily apparent on standard-of-care MR images. Here, we report integrating these images into a prediction model of tumor growth with radiation therapy. Materials and Methods: To predict therapy response, a mathematical algorithm of tumor infiltration based on the classical reaction-diffusion equation was customized to create a map of tumoral density for every location in the brain from the presence of choline (Cho), a proliferation-derived metabolite and N-acetylaspartate (NAA), a neuronal metabolite. Input images are pre and mid-treatment (2 weeks) MRSI acquisitions. The model is initialized using the gradient of the Cho/NAA ratio in the vicinity of tumor (periphery) as a measure of tumor infiltration, and the absolute Cho level as a measure of actively proliferating tumor cells. Customized, per-patient model parameters are tested until simulated images of tumor growth derived from the pre-treatment MRSI datasets match mid-treatment imaging at 2 weeks. Customization is done by iteratively modifying model parameters to minimize the discrepancy between simulated and acquired images as judged by the squared sum differences of voxel intensities within the PTV. We hypothesize that this model will predict response at week 10, 4 weeks following completion of concurrent radiation and temozolomide. Results: Utility of the MRSI-based tumor growth model was investigated by comparing the net treatment gain against clinical observations in GBM patients. Model parameters of tumor proliferation plus effectiveness of therapy correlated with clinical outcomes being positive if progression free survival is longer than 6 months. Conclusion: MRSI could provide more accurate information the reaction-diffusion tumor growth model than conventional MR images, as it provides information about proliferation and infiltration directly and can be easily integrated with clinical observations to predict per-patient response to therapy.

Purpose: Acoustic neuroma is treated using highly conformal, framed-based radiosurgery. The gross target volume (GTV) is located near the radiosensitive brainstem. Currently, plan evaluation methods that consider GTV dose conformity and dose to organs at risk (OAR) do not account for setup uncertainty, thereby potentially giving a false sense of target coverage and OAR sparing. More realistic plan evaluations should consider delivery uncertainties. In this work, we introduce a new plan evaluation method based on probability and investigate the effect of translational and rotational setup errors. Methods: Clinically approved 18 acoustic neuroma treatment plans were studied retrospectively. For each treatment plan, 200 new dose distributions were re-calculated while introducing both translational (in all cardinal directions) and rotational errors (in all 3 orthogonal rotation angles). The magnitude of errors for each simulation was randomly generated from the Gaussian distribution with one standard deviation of 0.5 mm and 0.5 degree respectively. From the simulated dose distributions, an approximated probability density function of dose objective for both GTV and brainstem were calculated. The probability density function was then used to compute the probability of delivering prescribed dose to 90% of GTV for each patient. For brainstem, the probability of limiting maximum dose to 11Gy was calculated. These probabilistic metrics were then analyzed with respect to GTV volume, conformity index, and min and max dose to the GTV and brainstem. Results: Across all cases, the average probability of delivering prescribed dose to the 90% of GTV volume was 85±14%. Four patients showed > 50% risk of exceeding 11Gy limit to the brainstem. Conclusion: In highly conformal, single fraction radiosurgery, a small setup error can be critical. We found that the probabilistic assessment of plan objectives can be used to quantitatively assess true target coverage better than the conventional plan assessment metrics.

Purpose: Treatment planning for whole brain radiation treatment is technically a simple process but in practice it takes valuable clinical time of repetitive and tedious tasks. This report presents a method that automatically segments the relevant target and normal tissues and creates a treatment plan in only a few minutes after patient simulation. Methods: Segmentation is performed automatically through morphological operations on the soft tissue. The treatment plan is generated by searching a database of previous cases for patients with similar anatomy. In this search, each database case is ranked in terms of similarity using a customized metric designed for sensitivity by including only geometrical changes that affect the dose distribution. The database case with the best match is automatically modified to replace relevant patient info and isocenter position while maintaining original beam and MLC settings. Results: Fifteen patients were used to validate the method. In each of these cases the anatomy was accurately segmented to mean Dice coefficients of 0.970 ± 0.008 for the brain, 0.846 ± 0.009 for the eyes and 0.672 ± 0.111 for the lens as compared to clinical segmentations. Each case was then subsequently matched against a database of 70 validated treatment plans and themore » best matching plan (termed auto-planned), was compared retrospectively with the clinical plans in terms of brain coverage and maximum doses to critical structures. Maximum doses were reduced by a maximum of 20.809 Gy for the left eye (mean 3.533), by 13.352 (1.311) for the right eye, and by 27.471 (4.856), 25.218 (6.315) for the left and right lens. Time from simulation to auto-plan was 3-4 minutes. Conclusion: Automated database- based matching is an alternative to classical treatment planning that improves quality while providing a cost—effective solution to planning through modifying previous validated plans to match a current patient's anatomy.« less

Purpose: Severe CT artifacts can impair our ability to accurately calculate proton range thereby resulting in a clinically unacceptable treatment plan. In this work, we investigated a novel CT artifact correction method based on a coregistered MRI and investigated its ability to estimate CT HU and proton range in the presence of severe CT artifacts. Methods: The proposed method corrects corrupted CT data using a coregistered MRI to guide the mapping of CT values from a nearby artifact-free region. First patient MRI and CT images were registered using 3D deformable image registration software based on B-spline and mutual information. The CT slice with severe artifacts was selected as well as a nearby slice free of artifacts (e.g. 1cm away from the artifact). The two sets of paired MRI and CT images at different slice locations were further registered by applying 2D deformable image registration. Based on the artifact free paired MRI and CT images, a comprehensive geospatial analysis was performed to predict the correct CT HU of the CT image with severe artifact. For a proof of concept, a known artifact was introduced that changed the ground truth CT HU value up to 30% and up to 5cm error in proton range. The ability of the proposed method to recover the ground truth was quantified using a selected head and neck case. Results: A significant improvement in image quality was observed visually. Our proof of concept study showed that 90% of area that had 30% errors in CT HU was corrected to 3% of its ground truth value. Furthermore, the maximum proton range error up to 5cm was reduced to 4mm error. Conclusion: MRI based CT artifact correction method can improve CT image quality and proton range calculation for patients with severe CT artifacts.

Purpose: Generating brain tumor volume measurements in a reproducible and efficient manner is a difficult, yet necessary, component of response assessment. The purpose of this study was to adapt and validate a multilevel Fuzzy C-means clustering algorithms for ROI tumor segmentation to allow consistent volumetric comparisons at multiple sites. Methods: Preoperative contrast-enhanced T1W images from 37 glioblastoma cases were segmented using Fuzzy C-means clustering-based methods and compared to manually contoured volumes created by specialists. The same was done post-operatively, using subtracted images to eliminate intrinsically T1-hyperintense material (blood). Volume computations based on the MacDonald criteria were also used for comparison. Agreement and inter-rater variability between volumes produced with each method was assessed by determining the concordance correlation coefficient (CCC). Results: The MacDonald criteria method had poor agreement (CCC=0.350–0.972) with manual contouring pre- and postoperatively, while the proposed semi-automated methods exhibited very high agreement (CCC=0.839–0.995) with manual contouring before and after resection. Fuzzy C-means clustering with three classes was the most robust semi-automated method, showing better inter-rater agreement than the MacDonald criteria method for both pre- (CCC of 0.990 and 0.975, respectively) and post-operative cases (CCC of 0.983 and 0.576, respectively). Post-operative inter-rater agreement was significantly different between these methods (p < 0.001). Conclusion: The proposed semi-automated segmentation methods allow tumor volume measurements of MR images in a reliable and reproducible fashion necessary for measuring treatment response in glioblastoma patients in multicenter clinical trials.

Purpose: Identification of a metabolite signature that shows significant tumor cell infiltration into normal brain in regions that do not appear abnormal on standard MRI scans would be extremely useful for radiation oncologists to choose optimal regions of brain to treat, and to quantify response beyond the MacDonald criteria. We report on integration of high-resolution magnetic resonance spectroscopic imaging (HR-MRSI) with radiation dose escalation treatment planning to define and target regions at high risk for recurrence. Methods: We propose to supplement standard MRI with a special technique performed on an MRI scanner to measure the metabolite levels within defined volumes. Metabolite imaging was acquired using an advanced MRSI technique combining 3D echo-planar spectroscopic imaging (EPSI) with parallel acquisition (GRAPPA) using a multichannel head coil that allows acquisition of whole brain metabolite maps with 108 μl resolution in 12 minutes implemented on a 3T MR scanner. Elevation in the ratio of two metabolites, choline (Cho, elevated in proliferating high-grade gliomas) and N-acetyl aspartate (NAA, a normal neuronal metabolite), was used to image infiltrating high-grade glioma cells in vivo. Results: The metabolite images were co-registered with standard contrast-enhanced T1-weighted MR images using in-house registration software and imported into the treatment-planning system. Regionsmore » with tumor infiltration are identified on the metabolic images and used to create adaptive IMRT plans that deliver a standard dose of 60 Gy to the standard target volume and an escalated dose of 75 Gy (or higher) to the most suspicious regions, identified as areas with elevated Cho/NAA ratio. Conclusion: We have implemented a state-of-the-art HR-MRSI technology that can generate metabolite maps of the entire brain in a clinically acceptable scan time, coupled with introduction of an imaging co-registration/ analysis program that combines MRSI data with standard imaging studies in a clinically useful fashion.« less

Purpose: Imaging biomarkers are crucial in managing treatment options for cancer patients. They are extremely powerful tools since they allow personalized treatment assessment early during therapy by using repeated imaging to detect and quantify tumor response. Currently, treatment response assessment from consecutive imaging is measured by simple global measures that do not capture a tumor's heterogeneous response. The authors present an automated, multivoxel metric that groups voxels into clusters of changes for a local definition of radiation treatment efficiency from multiple PET imaging studies acquired at different time periods for assessing therapeutic response. Methods: The algorithm employs level-set mathematics to extract changing features to classify voxels into response patterns. First, pretreatment and post-treatment PET images were aligned using a deformable registration to correct for posture and soft tissue changes. The detailed mapping was modeled by free form deformations B-spline optimized using the limited memory L-BFGS algorithm. The posture-corrected datasets are then subtracted to produce an image of molecular changes embedded with noise. Once images were aligned and subtracted, a segmentation algorithm combining the concepts of voxel and distance-based techniques classified voxels into patterns of signal reduction or enhancement. Although signal reduction is evidence of successful treatment, signal-enhancing regions are an indication of treatment failure. For an in depth analysis of potential treatment errors, patterns of signal enhancement were correlated with the radiation treatment dose and anatomical structures from the treatment plan using image registration methods. Results: The algorithm was retrospectively applied to PET/CT and radiotherapy (RT) oncology data from an NCI-sponsored clinical trial (81 clinical cases from RTOG 0522 Trial) for combined drug and radiation therapy in head and neck carcinomas. This clinical trial dataset presented a realistic environment for implementing and validating our algorithm to correlate local response as observed in serial PET with delivered dose. The technique was instrumental in detecting geographical and segmentation misses on the actual clinical cases by providing accurate voxel-by-voxel analysis of metabolic changes. Results of the level-set based clustering algorithm are saved as a detailed report of enhancing/nonenhancing regions and their location, and can be further displayed as a colorwash laid over the original anatomy for in depth analysis. Conclusions: The automated technique was instrumental in analyzing treatment response in the clinical cases and provided an useful tool for accurate, outcome-based response assessment of the radiation treatment process. The developed method is general and should be extendable to other high-resolution diagnostic imaging with minor modifications.

Purpose: Repeated imaging is an extremely powerful tool in current radiotherapy practice since it allows advanced tumordetection and personalized treatment assessment by quantify tumor response. Change detection algorithms have been developed for remote sensingimages to mathematically quantify relevant modifications occurring between datasets of the same subject acquired at different times. We propose usage of change detectors in radiotherapy for an automated quantification of clinical changes occurring in repeated imaging.Methods: We explore usage of the Kullback‐Leiber divergence as indicator of tumor change and quantification of treatment response. The Kullbach‐Leiber divergence uses the likelihood theory to measures the distance between two statistical distributions and thus does not assume consistency in imaging. By it's general nature, it can accommodate the presence of noise and variations in imaging acquisition parameters that usually hinder automated identification of clinically‐relevant features. Results: In a comparison of simple difference maps and the Kullbach‐Leiber divergence operator, the difference maps were affected by noise and did not consistently detect changes of low intensity. In contrast, the proposed operator discerned noise by considering regional statistics around each voxel, and marked both regions with low and high contrast changes. Conclusions: Statistical comparison through Kullback‐Leiber divergence provides a reliable means to automatically quantify changes in repeated radiotherapyimaging

Purpose: Anatomical segmentation remains the most time‐consuming task in radiotherapy planning. Atlas‐based segmentation methods using deformable registration emerged as a practical approach to automate the process for head and neck anatomy, but it poses significant challenges in thoracic and pelvic regions where large inter‐patient variations are present. We present an approach combining diffeomorphic inter‐modality deformable registration with level‐set based refinement for an accurate atlas‐ based segmentation of CT scans. Methods: Our automated approach involves a multi‐atlas matching stage followed by a per‐organ refinement. At the atlas matching stage, multiple atlases are matched to the patient anatomy to determine a rough segmentation using a new non‐parametric multi‐modality diffeomorphic deformable registration algorithm able to cope with differences in contrast, bowel filling, and artifacts. A probabilistic map of organ shape and locations is constructed from each atlas mapping using the STAPLE algorithm, followed by a refinement procedure using variants of level‐set algorithms customized for each organ. Validation was performed by comparing results against clinical manual segmentations using the Dice coefficientResults: Dice comparison with manual segmentation measured on 20 clinical cases ranged 0.89–0.92 for liver, 0.91–0.93 for kidneys, 0.94–0.96 for spleen 0.87–0.92 for heart and 0.87–0.92 for the spinal cord. The algorithm is fully automated, running as a software service in the background without any user interaction Conclusions: Automated segmentation accuracy for abdominal and thoracic regions is affected without a multi‐atlas and per‐structure refinement strategy. The process significantly decreases segmentation times improving clinical workflow and efficiency as segmentations of normal tissue are automatically performed immediately after image acquisition.

Purpose: To develop an automatic and accurate technique for concomitant segmentations and registration of liveranatomy using SPECT and CTimages for unsealed sourceradiotherapy.Method and Materials: The link between segmentation and registration is given by the using the level set of a liver segmentation into the registration process. In the combined approach, the liver is automatically segmented from the CTimage by evolving an initial seed with a level set until it locks to the liver's border as observed in the CTimages. The time‐crossing map of the level set is then used to match gradients in the SPECTimage to the level set by using a data structure containing the signed distance values at a small band of neighboring pixels. Results: The technique was applied to three cases of metastatic liver disease treated with unsealed source therapy. Results indicated that the speed map of the level set plays an importance role in obtainng an accurate registration and produce a segmentation that is superior in registration time and accuracy over manual segmentation or the standard registration approach using mutual information. Accuracy measured with the convergence analysis method was of less than 0.5 mm rotation and 1 degree rotation. Conclusion: With the proposed combined segmentation‐registration technique, the uncertainty of soft‐tissue target localization could be greatly reduced ensuring accurate therapy assessment to be precisely delivered as planned. The combined all‐in‐one approach is automated and provides excellent accuracy over manual segmentation and mutual information approaches.

Purpose: We design and implement linear programming based IMRT treatment plan generation technology that effectively and efficiently optimizes beam geometry as well as beamlet intensities. Method and Materials: To overcome the limitations of nonlinear and/or mixed integer programs associated with dose‐volume constraints, we adapted the conditional value‐at‐risk constraints (C‐VaR constraints) which are popular and commonly used in financial engineering. The use of C‐Var constraints has significant computational advantages as they can be handled using linear programming models, which are easily and efficiently solved. However, the parameters controlling the C‐Var constraints have to be chosen carefully to get an accurate approximation of the dose‐volume constraints. The treatment plan generation technology, which is based on an optimizationmodel using C‐Var constraints, embeds an effective and efficient parameter search scheme to ensure high‐quality approximations of the dose‐volume constraints. We will report on application of the system to three patient cases: head‐and‐neck cancer, prostate cancer and pediatric braintumor.Results: The utilization of our system returned optimal treatment plans for a variety of cases. Optimal beam selection, as created by our software, generally improves the coverage and conformity of the treatment plan without sacrificing normal tissue sparing. There are generally limits to how much the solution is improved by adding additional beams which vary from case to case. Calculational time was generally a fraction of the time that mixed‐integer solutions would take. Conclusion: Developed treatment plan generation technology optimizes both beam geometry and beamlet intensities in a short amount of time. The technology is fully automated and generates several high‐quality treatment plans satisfying the provided minimum requirements in a single invocation and without human guidance.

Purpose: The in‐vivo imaging of tumor biology through positron emission tomography(PET) scans has become an invaluable tool for therapy response assessment since it provides tumor response data at early treatment stages. However, imaging in‐vivo small molecular changes in tissues subject to respiratory motion is infeasible as the PET tracer is smeared during the acquisition process introducing errors up to 75% associated with degraded tracer activity. The purpose of this study was to develop a motion detectionsystem that quantifies respiratory changes using spatio‐temporal Bspline deformable image registration for analysis of motion‐degraded PET datasets. Method and Materials: A clinical software tool implemented the Bspline deformable image registration method. Algorithm output is a motion‐corrected 4D PETimage containing only treatment‐induced changes. The motion‐free dataset can be summed to provide a simple 3D dataset that is used to for therapy evaluation using the standard tools developed for motion‐free treatment sites. Results: Capabilities of motion correction algorithm were established in studies on tumor motion in dynamic phantoms using checkerboard and surface distance tools and in patients using the convergence analysis method. The BSpline model was able to reproduce respiratory motion with a maximum error of 3 mm on five clinical cases of lung lesions. Conclusion: The integrated motion‐correction 4D PETsystem will provide a standardized protocol for characterization of tumor response using a combination of conventional and motion‐sensitive measures. The algorithm discerns between respiratory and treatment induced changes and thus, enables individual tumor response and evaluation of anti‐angiogenic therapies in images of thoracic tumors affected by respiration.

Purpose: To report on a tool that automatically extracts the intra‐fraction motion from a 4D CT dataset and integrates the information with treatment planning which allows accurate calculation of delivered dose in each fraction and in the presence of respiratory motion. Method: For this project, we have integrated the ITK/VTK libraries with a commercial planning system, Eclipse (Varian Medical Systems, Palo Alto, CA). The 4D registration tool provided a smooth description of the respiratory motion and was used to map contours delineated in one phase to all phases of a 4D CT dataset. Resulting segmentation was saved in DICOM format and imported together with the associated datasets in Eclipse. Dose calculation and dose‐volume histograms were calculated directly for each phase of the dataset using the original beam fluences. The dose in each phase can be added or subtracted directly in Eclipse to provide the dose summation over the whole respiratory cycle or to inspect the dose differences between phases produced by motion. Computation of 4D DVHs for static, serial organ such as the spinal cord is possible directly with Eclipse. For moving parallel organs such as the lung, we provide an independent DICOM‐compliant tool to sum‐up individual dose matrixes taking into account organ motion. Results: Dose differences between the inhalation phase (used for planning) and the exhalation phase can be easily assessed using the plan evaluation tool. By summing treatment plans for each phase, comparison of the planned and delivered dose distributions can be assessed in treatment planning.Conclusion: Individual dose verification of treatment plans in presence of intra‐fraction motion is possible using available commercial treatment planning systems.

Seven children with advanced nasopharyngeal carcinoma younger than 20 years of age diagnosed between 1975 and 1986 (inclusive) were treated with a uniform adjuvant chemotherapy regimen, which consisted of vincristine (1.5 mg/m2; day 1), doxorubicin (45 mg/m2; day 1), 5-fluorouracil (8 mg/kg; days 1 through 5), and cyclophosphamide (7 mg/kg; days 1 through 5). This combination chemotherapy was given for 12 to 24 months after completion of radiation therapy. The radiation doses to the primary sites ranged from 6000 cGy to a maximum of 6800 cGy. The radiation doses for neck prophylaxis ranged from 4500 cGy to a total of 5000 cGy. Involved sites were irradiated to at least an additional boost of 1000 cGy. One patient had an external dose 6000 cGy to the primary site boosted with brachytherapy of 3000 cGy at the surface of an ovoid. After chemotherapy myelosuppression occurred in all patients and was tolerable. All seven patients are surviving, six disease-free, for 22 months to 12 years (median, 4 years). This study suggests that the combination of radiation therapy and chemotherapy as used here has acceptable toxicity and is effective and further suggests that children with nasopharyngeal carcinoma, even in its advanced stage at diagnosis, may be curable.

In a prospective study of squamous head and neck cancer, the pretreatment peripheral blood of 125 patients was examined for lymphocyte subclass and in vitro immunologic function. After 4 years of followup, 49 recurrences of disease were observed. Lymphocytes from patients with recurrent disease showed elevated interleukin-2 (IL-2) production and a tendency towards increased response to mitogens in comparison to those without recurrence. When disease-free survival is analyzed on the basis of IL-2 levels, patients with high relative IL-2 synthesis (21) had a 40% poorer prognosis than patients with low relative IL-2 levels (4). The difference is significant at a P value of 0.02. Since IL-2 synthesis occurs with antigenic stimulation, it is postulated that patients who have a high IL-2 synthesis in their pretreatment lymphocytes may have had prior stimulation by circulating tumor antigen. Such immune response by the host may be successful in destroying the antigenic tumor cells but may leave the undifferentiated, less antigenic tumor cells to grow and metastasize. Thus, elevated IL-2 synthesis in pretreatment lymphocytes predicts a poorer prognosis. S solid tumors, usually has a predictable biologic and clinical behavior. This carcinoma presumably begins with a precancerous condition and progresses to an invasive state manifested by local and regional involvement, followed subsequently by dissemination to distant organs.' In the initial stages of disease, surgery and/or radiation are curative methods of treatment. For advanced stages, these two modalities of treatment are insufficient to control the disease. The role of adjunctive chemotherapy is presently under close investigation. Many important factors have been observed to affect the outcome of treatment, namely, anatomic origin of the cancer, extent of the disease as determined by staging, biologic behaviors expressed either by the tempo of disease progression clinically, or invasiveness and metastatic potential as determined by resectability and early recurrence, age of the