Image-Guided Radiotherapy Shift Data



Residual setup errors towards the heart after image guidance linked with poorer survival in lung cancer patients: do we need stricter IGRT protocols?Residual setup errors towards the heart linked to poorer survivalSummaryImage-guided radiotherapy (IGRT) is widely used but data providing evidence of its direct effect on patient outcome is scarce. We related residual patient setup errors from IGRT data to overall survival for NSCLC. The direction of the residual errors was found to be significant, with patients with residual shifts towards the heart having significantly worse overall survival than those with shifts away from the heart. This result was independently validated in an oesophageal cancer cohort.AbstractBackground and Purpose: Image-guided radiotherapy (IGRT) is widely used but data directly relating setup errors to patient outcome is scarce. This study investigates the relationship between residual IGRT shifts and overall patient survival, and uses the observed relations to identify structures sensitive to radiation dose.Materials and Methods: Residual shift data for 780 NSCLC patients was summarised for each patient over the course of treatment by determining the mean shifts, standard deviations and the vector shift in the direction of the heart. These variables were related to overall survival, and significant variables were used to produce Kaplan-Meier plots of survival. The effect of shift directionality was studied by splitting the cohort into left, right, anterior, posterior, superior and inferior groups, and by analysing the vector shift in the direction of the heart. The observed relationship was independently validated in an oesophageal cancer cohort (n = 177).Results: The shift data showed strong associations with survival. Left and right cohorts showed opposite directional shift effects, suggesting shifts towards the mediastinum have a negative effect on survival. Projection of the vector shift in the direction of the heart showed that patients with a residual shift towards the heart have significantly worse overall survival (p = 0.007, hazard ratio 1.091). The same effect was observed in the oesophageal cancer cohort (p = 0.041, hazard ratio 1.164).Conclusions: Residual shift metrics derived from IGRT data can categorise both NSCLC and oesophageal cancer patients into populations with significantly different survival times based on the size of the residual shift in the direction of the heart, thus providing evidence of the importance of using strict IGRT protocols to spare organs at risk and at the same time highlighting the heart as a dose sensitive organ.IntroductionLung cancer is one of the most commonly occurring cancers worldwide ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "URL" : "", "accessed" : { "date-parts" : [ [ "2017", "11", "6" ] ] }, "author" : [ { "dropping-particle" : "", "family" : "Cancer Research UK", "given" : "", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "id" : "ITEM-1", "issued" : { "date-parts" : [ [ "0" ] ] }, "title" : "Cancer Statistics - Worldwide Cancer", "type" : "webpage" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[1]", "plainTextFormattedCitation" : "[1]", "previouslyFormattedCitation" : "[1]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[1], with approximately 46,000 new cases being diagnosed each year in the UK alone ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "URL" : "", "accessed" : { "date-parts" : [ [ "2017", "3", "23" ] ] }, "author" : [ { "dropping-particle" : "", "family" : "Cancer Research UK", "given" : "", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "id" : "ITEM-1", "issued" : { "date-parts" : [ [ "0" ] ] }, "title" : "About Cancer - Lung Cancer", "type" : "webpage" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[2]", "plainTextFormattedCitation" : "[2]", "previouslyFormattedCitation" : "[2]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[2]. Lung cancer patients receive one or a combination of surgery, radiotherapy and chemotherapy, with radiotherapy playing an important role for all tumour stages. Radiotherapy plans are numerically optimized on Computed Tomography (CT) images, acquired with the patient in their treatment position. At each treatment fraction this position must be replicated as accurately as possible as deviations from the planned position will result in a different dose distribution being delivered than which was planned, which could affect the probability of tumour control and normal tissue complications. Previous studies have confirmed that the quality of the delivered radiotherapy can greatly affect patient outcomes ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1200/JCO.2009.27.4498", "ISBN" : "1527-7755 (Electronic)\\r0732-183X (Linking)", "ISSN" : "0732183X", "PMID" : "20479390", "abstract" : "PURPOSE: To report the impact of radiotherapy quality on outcome in a large international phase III trial evaluating radiotherapy with concurrent cisplatin plus tirapazamine for advanced head and neck cancer. PATIENTS AND METHODS: The protocol required interventional review of radiotherapy plans by the Quality Assurance Review Center (QARC). All plans and radiotherapy documentation underwent post-treatment review by the Trial Management Committee (TMC) for protocol compliance. Secondary review of noncompliant plans for predicted impact on tumor control was performed. Factors associated with poor protocol compliance were studied, and outcome data were analyzed in relation to protocol compliance and radiotherapy quality. RESULTS: At TMC review, 25.4% of the patients had noncompliant plans but none in which QARC-recommended changes had been made. At secondary review, 47% of noncompliant plans (12% overall) had deficiencies with a predicted major adverse impact on tumor control. Major deficiencies were unrelated to tumor subsite or to T or N stage (if N+), but were highly correlated with number of patients enrolled at the treatment center (< five patients, 29.8%; > or = 20 patients, 5.4%; P < .001). In patients who received at least 60 Gy, those with major deficiencies in their treatment plans (n = 87) had a markedly inferior outcome compared with those whose treatment was initially protocol compliant (n = 502): -2 years overall survival, 50% v 70%; hazard ratio (HR), 1.99; P < .001; and 2 years freedom from locoregional failure, 54% v 78%; HR, 2.37; P < .001, respectively. CONCLUSION: These results demonstrate the critical importance of radiotherapy quality on outcome of chemoradiotherapy in head and neck cancer. Centers treating only a few patients are the major source of quality problems.", "author" : [ { "dropping-particle" : "", "family" : "Peters", "given" : "Lester J.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "O'Sullivan", "given" : "Brian", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Giralt", "given" : "Jordi", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Fitzgerald", "given" : "Thomas J.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Trotti", "given" : "Andy", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Bernier", "given" : "Jacques", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Bourhis", "given" : "Jean", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Yuen", "given" : "Kally", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Fisher", "given" : "Richard", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Rischin", "given" : "Danny", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Journal of Clinical Oncology", "id" : "ITEM-1", "issue" : "18", "issued" : { "date-parts" : [ [ "2010" ] ] }, "page" : "2996-3001", "title" : "Critical impact of radiotherapy protocol compliance and quality in the treatment of advanced head and neck cancer: Results from TROG 02.02", "type" : "article-journal", "volume" : "28" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[3]", "plainTextFormattedCitation" : "[3]", "previouslyFormattedCitation" : "[3]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[3] and the negative effects that deviations can have on tumour control ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/j.ijrobp.2004.11.032", "ISBN" : "0360-3016 (Print)\\r0360-3016 (Linking)", "ISSN" : "03603016", "PMID" : "15989996", "abstract" : "Purpose: To retrospectively test the hypothesis that rectal distension on the planning computed tomography (CT) scan is associated with an increased risk of biochemical and local failure among patients irradiated for prostate carcinoma when a daily repositioning technique based on direct prostate-organ localization is not used. Methods and Materials: This study included 127 patients who received definitive three-dimensional conformal radiotherapy for prostate cancer to a total dose of 78 Gy at The University of Texas M. D. Anderson Cancer Center. Rectal distension was assessed by calculation of the average cross-sectional rectal area (CSA; defined as the rectal volume divided by length) and measuring three rectal diameters on the planning CT. The impact of rectal distension on biochemical control, 2-year prostate biopsy results, and incidence of Grade 2 or greater late rectal bleeding was assessed. Results: The incidence of biochemical failure was significantly higher among patients with distended rectums (CSA >11.2 cm2) on the planning CT scan (p = 0.0009, log-rank test). Multivariate analysis indicates that rectal distension and high-risk disease are independent risk factors for biochemical failure, with hazard ratios of 3.89 (95% C.I. 1.58 to 9.56, p = 0.003) and 2.45 (95% C.I. 1.18 to 5.08, p = 0.016), respectively. The probability of residual tumor without evidence of radiation treatment (as scored by the pathologist) increased significantly with rectal distension (p = 0.010, logistic analysis), and a lower incidence of Grade 2 or greater late rectal bleeding within 2 years was simultaneously observed with higher CSA values (p = 0.031, logistic analysis). Conclusions: We found strong evidence that rectal distension on the treatment-planning CT scan decreased the probability of biochemical control, local control, and rectal toxicity in patients who were treated without daily image-guided prostate localization, presumably because of geographic misses. Therefore, an empty rectum is warranted at the time of simulation. These results also emphasize the need for image-guided radiotherapy to improve local control in irradiating prostate cancer. \u00a9 2005 Elsevier Inc.", "author" : [ { "dropping-particle" : "", "family" : "Crevoisier", "given" : "Renaud", "non-dropping-particle" : "De", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Tucker", "given" : "Susan L.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Dong", "given" : "Lei", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Mohan", "given" : "Radhe", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Cheung", "given" : "Rex", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Cox", "given" : "James D.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Kuban", "given" : "Deborah A.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "International Journal of Radiation Oncology Biology Physics", "id" : "ITEM-1", "issue" : "4", "issued" : { "date-parts" : [ [ "2005" ] ] }, "page" : "965-973", "title" : "Increased risk of biochemical and local failure in patients with distended rectum on the planning CT for prostate cancer radiotherapy", "type" : "article-journal", "volume" : "62" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[4]", "plainTextFormattedCitation" : "[4]", "previouslyFormattedCitation" : "[4]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[4].Image guided radiotherapy (IGRT) has been widely implemented over the last decades to aid patient positioning. Typically using integrated planar kV or MV radiographs, or cone-beam CT (CBCT), patients are imaged on the treatment machine prior to irradiation and their position and anatomy are compared to that at planning ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/S1470-2045(06)70904-4", "abstract" : "Technological advances have greatly enhanced the specialty of radiation oncology by allowing more healthy tissue to be spared for the same or better tumour coverage. Developments in medical imaging are integral to radiation oncology, both for design of treatment plans and to localise the target for precise administration of radiation. At planning, defi nition of the tumour and healthy tissue is based on CT, augmented frequently with MRI and PET. At treatment, three-dimensional soft-tissue imaging can also be used to localise the target and tumour motion can be tracked with fl uoroscopic imaging of radio-opaque markers implanted in or near the tumour. These developments allow changes in tumour position, size, and shape that take place during radiotherapy to be measured and accounted for to boost geometric accuracy and precision of radiation delivery. Image-guided treatment also enhances uniformity in doses administered in a population of patients, thus improving our ability to measure the eff ect of dosimetric and non-dosimetric factors on tumour and healthy tissue outcomes in clinical trials. Increased precision and accuracy of radiotherapy are expected to augment tumour control, reduce incidence and severity of toxic eff ects after radiotherapy, and facilitate development of more effi cient shorter schedules than currently available.", "author" : [ { "dropping-particle" : "", "family" : "Dawson", "given" : "Laura A", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Sharpe", "given" : "Michael B", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Lancet Oncology", "id" : "ITEM-1", "issue" : "10", "issued" : { "date-parts" : [ [ "2006" ] ] }, "page" : "848-858", "title" : "Image-guided radiotherapy : rationale , benefits , and limitations", "type" : "article-journal", "volume" : "7" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[5]", "plainTextFormattedCitation" : "[5]", "previouslyFormattedCitation" : "[5]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[5]. Numerous studies have demonstrated the advantages of using IGRT for patient positioning ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/j.ijrobp.2008.08.059", "ISBN" : "0360-3016", "ISSN" : "03603016", "PMID" : "19095368", "abstract" : "Purpose To determine the geometric accuracy of lung cancer radiotherapy using daily volumetric, cone-beam CT (CBCT) image guidance and online couch position adjustment. Methods and Materials Initial setup accuracy using localization CBCT was analyzed in three lung cancer patient cohorts. The first (n = 19) involved patients with early-stage non\u2013small-cell lung cancer (NSCLC) treated using stereotactic body radiotherapy (SBRT). The second (n = 48) and third groups (n = 20) involved patients with locally advanced NSCLC adjusted with manual and remote-controlled couch adjustment, respectively. For each group, the couch position was adjusted when positional discrepancies exceeded \u00b13 mm in any direction, with the remote-controlled couch correcting all three directions simultaneously. Adjustment accuracy was verified with a second CBCT. Population-based setup margins were derived from systematic (\u03a3) and random (\u03c3) positional errors for each group. Results Localization imaging demonstrates that 3D positioning errors exceeding 5 mm occur in 54.5% of all delivered fractions. CBCT reduces these errors; post-correction \u03a3 and \u03c3 ranged from 1.2 to 1.9 mm for Group 1, with 82% of all fractions within \u00b13 mm. For Group 2, \u03a3 and \u03c3 ranged between 0.8 and 1.8 mm, with 76% of all treatment fractions within \u00b13 mm. For Group 3, the remote-controlled couch raised this to 84%, and \u03a3 and \u03c3 were reduced to 0.4 to 1.7 mm. For each group, the postcorrection setup margins were 4 to 6 mm, 3 to 4 mm, and 2 to 3 mm, respectively. Conclusions Using IGRT, high geometric accuracy is achievable for NSCLC patients, potentially leading to reduced PTV margins, improved outcomes and empowering adaptive radiation therapy for lung cancer.", "author" : [ { "dropping-particle" : "", "family" : "Bissonnette", "given" : "Jean-Pierre", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Purdie", "given" : "Thomas G.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Higgins", "given" : "Jane A.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Li", "given" : "Winnie", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Bezjak", "given" : "Andrea", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "International Journal of Radiation Oncology*Biology*Physics", "id" : "ITEM-1", "issue" : "3", "issued" : { "date-parts" : [ [ "2009" ] ] }, "page" : "927-934", "title" : "Cone-Beam Computed Tomographic Image Guidance for Lung Cancer Radiation Therapy", "type" : "article-journal", "volume" : "73" }, "uris" : [ "" ] }, { "id" : "ITEM-2", "itemData" : { "DOI" : "10.1016/j.meddos.2009.09.003", "author" : [ { "dropping-particle" : "", "family" : "Li", "given" : "Winnie", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Moseley", "given" : "Douglas J", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Bissonnette", "given" : "Jean-Pierre", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Purdie", "given" : "Thomas G.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Bezjak", "given" : "Andrea", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Jaffray", "given" : "David A.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Medical Dosimetry", "id" : "ITEM-2", "issue" : "4", "issued" : { "date-parts" : [ [ "2010" ] ] }, "page" : "287-296", "title" : "Setup Reproducibility for Thoracic and Upper Gastrointestinal Radiation Therapy: Influence of Immobilization Method and On-Line Cone-Beam CT Guidance", "type" : "article-journal", "volume" : "35" }, "uris" : [ "" ] }, { "id" : "ITEM-3", "itemData" : { "DOI" : "10.1016/j.radonc.2011.07.027", "ISSN" : "1879-0887", "PMID" : "21872956", "abstract" : "BACKGROUND AND PURPOSE: To evaluate residual patient setup errors and daily dose variations of different less-than-daily image guidance (IG) strategies in the delivery of external beam radiotherapy for esophageal cancer.\n\nMATERIAL AND METHODS: Daily image-guided setup data for 25 consecutive esophageal cancer patients treated with helical tomotherapy were evaluated. Seven less-than-daily IG strategies with different imaging frequencies were simulated. For each IG strategy, the daily residual setup errors were calculated. Using TomoTherapy Planned Adaptive software, daily dose variations to the clinical target volume, heart, and lungs were evaluated in five representative patients.\n\nRESULTS: With 0% (60%) IG frequency, the margins required for adequate coverage of the clinical target volume were 13 mm (10 mm), 14 mm (11 mm), and 5 mm (5 mm) in the left-right, superior-inferior, and anterior-posterior directions, respectively. Even with 60% IG frequency, 10% of the fractions had more than 10% decrease in the dose level covering 95% of the target, and 14% and 13% of the fractions had more than 10% increase in total lung volume receiving at least 0.8 Gy per fraction, and heart volume receiving at least 1.2 Gy per fraction, respectively.\n\nCONCLUSION: Substantial residual setup errors would occur for treatment fractions without IG even if the most frequent less-than-daily IG strategy was to be used, which could lead to significant daily dose variations for the target volume and adjacent normal tissues. Daily image guidance is recommended throughout the course of treatment in conformal radiotherapy for esophageal cancer.", "author" : [ { "dropping-particle" : "", "family" : "Han", "given" : "Chunhui", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Schiffner", "given" : "Daniel C", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Schultheiss", "given" : "Timothy E", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Chen", "given" : "Yi-Jen", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Liu", "given" : "An", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Wong", "given" : "Jeffrey Y C", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Radiotherapy and oncology : journal of the European Society for Therapeutic Radiology and Oncology", "id" : "ITEM-3", "issue" : "2", "issued" : { "date-parts" : [ [ "2012", "2" ] ] }, "page" : "309-14", "publisher" : "Elsevier Ireland Ltd", "title" : "Residual setup errors and dose variations with less-than-daily image guided patient setup in external beam radiotherapy for esophageal cancer.", "type" : "article-journal", "volume" : "102" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[6\u20138]", "plainTextFormattedCitation" : "[6\u20138]", "previouslyFormattedCitation" : "[6\u20138]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[6–8], reporting superior geometric and dosimetric conformance to the plan. As such, IGRT is now routinely used for the correction of patient setup errors ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/S0360-3016(01)01624-8", "author" : [ { "dropping-particle" : "", "family" : "Boer", "given" : "Hans C J", "non-dropping-particle" : "de", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Heijmen", "given" : "Ben J M", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "International Journal of Radiation Oncology*Biology*Physics", "id" : "ITEM-1", "issue" : "5", "issued" : { "date-parts" : [ [ "2001" ] ] }, "page" : "1350-1365", "title" : "A Protocol for the Reduction of Systematic Patient Setup Errors with Minimal Portal Imaging Workload", "type" : "article-journal", "volume" : "50" }, "uris" : [ "" ] }, { "id" : "ITEM-2", "itemData" : { "DOI" : "10.1016/J.IJROBP.2006.11.050", "ISSN" : "0360-3016", "author" : [ { "dropping-particle" : "", "family" : "Boer", "given" : "Hans C J", "non-dropping-particle" : "de", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Heijmen", "given" : "Ben J M", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "International Journal of Radiation Oncology*Biology*Physics", "id" : "ITEM-2", "issue" : "5", "issued" : { "date-parts" : [ [ "2007", "4", "1" ] ] }, "page" : "1586-1595", "publisher" : "Elsevier", "title" : "eNAL: An Extension of the NAL Setup Correction Protocol for Effective Use of Weekly Follow-up Measurements", "type" : "article-journal", "volume" : "67" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[9,10]", "plainTextFormattedCitation" : "[9,10]", "previouslyFormattedCitation" : "[9,10]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[9,10], has facilitated the use of hypofractionation by providing confidence in the patient setup, and also finds application in adaptive radiotherapy strategies ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/j.ijrobp.2007.05.046", "author" : [ { "dropping-particle" : "", "family" : "Nijkamp", "given" : "Jasper", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Pos", "given" : "Floris J", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Nuver", "given" : "Tonnis T", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Jong", "given" : "Rianne", "non-dropping-particle" : "de", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Remeijer", "given" : "Peter", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Sonke", "given" : "Jan-Jakob", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "V", "family" : "Lebesque", "given" : "Joos", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "International Journal of Radiation Oncology*Biology*Physics", "id" : "ITEM-1", "issue" : "1", "issued" : { "date-parts" : [ [ "2008" ] ] }, "page" : "75-82", "title" : "Adaptive Radiotherapy for Prostate Cancer using Kilovoltage Cone-Beam Computed Tomography: First Clinical Results", "type" : "article-journal", "volume" : "70" }, "uris" : [ "" ] }, { "id" : "ITEM-2", "itemData" : { "DOI" : "10.1016/j.semradonc.2009.11.003", "ISSN" : "10534296", "author" : [ { "dropping-particle" : "", "family" : "Sonke", "given" : "Jan-Jakob", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Belderbos", "given" : "Jos\u00e9", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Seminars in Radiation Oncology", "id" : "ITEM-2", "issue" : "2", "issued" : { "date-parts" : [ [ "2010", "4" ] ] }, "page" : "94-106", "publisher" : "Elsevier Inc.", "title" : "Adaptive Radiotherapy for Lung Cancer", "type" : "article-journal", "volume" : "20" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[11,12]", "plainTextFormattedCitation" : "[11,12]", "previouslyFormattedCitation" : "[11,12]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[11,12]. Much of the evidence base for IGRT, as mentioned above, relies upon surrogate geometric and dosimetric outcomes with the implicit assumption that better conformance to the planned treatment correlates with improved clinical outcome. The only study directly relating clinical outcome to the use of IGRT that the authors’ are aware of is the work of Zelefsky et al ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/j.ijrobp.2011.11.047", "ISBN" : "0360-3016\\r1879-355X", "ISSN" : "03603016", "PMID" : "22330997", "abstract" : "Purpose: To compare toxicity profiles and biochemical tumor control outcomes between patients treated with high-dose image-guided radiotherapy (IGRT) and high-dose intensity-modulated radiotherapy (IMRT) for clinically localized prostate cancer. Materials and Methods: Between 2008 and 2009, 186 patients with prostate cancer were treated with IGRT to a dose of 86.4 Gy with daily correction of the target position based on kilovoltage imaging of implanted prostatic fiducial markers. This group of patients was retrospectively compared with a similar cohort of 190 patients who were treated between 2006 and 2007 with IMRT to the same prescription dose without, however, implanted fiducial markers in place (non-IGRT). The median follow-up time was 2.8 years (range, 2-6 years). Results: A significant reduction in late urinary toxicity was observed for IGRT patients compared with the non-IGRT patients. The 3-year likelihood of grade 2 and higher urinary toxicity for the IGRT and non-IGRT cohorts were 10.4% and 20.0%, respectively (p = 0.02). Multivariate analysis identifying predictors for grade 2 or higher late urinary toxicity demonstrated that, in addition to the baseline Internatinoal Prostate Symptom Score, IGRT was associated with significantly less late urinary toxicity compared with non-IGRT. The incidence of grade 2 and higher rectal toxicity was low for both treatment groups (1.0% and 1.6%, respectively; p = 0.81). No differences in prostate-specific antigen relapse-free survival outcomes were observed for low- and intermediate-risk patients when treated with IGRT and non-IGRT. For high-risk patients, a significant improvement was observed at 3 years for patients treated with IGRT compared with non-IGRT. Conclusions: IGRT is associated with an improvement in biochemical tumor control among high-risk patients and a lower rate of late urinary toxicity compared with high-dose IMRT. These data suggest that, for definitive radiotherapy, the placement of fiducial markers and daily tracking of target positioning may represent the preferred mode of external-beam radiotherapy delivery for the treatment of prostate cancer. \u00a9 2012 Elsevier Inc. All rights reserved.", "author" : [ { "dropping-particle" : "", "family" : "Zelefsky", "given" : "Michael J.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Kollmeier", "given" : "Marisa", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Cox", "given" : "Brett", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Fidaleo", "given" : "Anthony", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Sperling", "given" : "Dahlia", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Pei", "given" : "Xin", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Carver", "given" : "Brett", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Coleman", "given" : "Jonathan", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Lovelock", "given" : "Michael", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Hunt", "given" : "Margie", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "International Journal of Radiation Oncology Biology Physics", "id" : "ITEM-1", "issue" : "1", "issued" : { "date-parts" : [ [ "2012" ] ] }, "page" : "125-129", "publisher" : "Elsevier Inc", "title" : "Improved clinical outcomes with high-dose image guided radiotherapy compared with non-IGRT for the treatment of clinically localized prostate cancer", "type" : "article-journal", "volume" : "84" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[13]", "plainTextFormattedCitation" : "[13]", "previouslyFormattedCitation" : "[13]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[13] in prostate cancer. They showed that use of daily fiducial marker based set-up error correction improved biochemical tumour control and reduced incidence of late urinary toxicity, as compared to a cohort treated without IGRT. As discussed by Bujold et al ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/j.semradonc.2011.09.001", "ISBN" : "1053-4296", "ISSN" : "10534296", "PMID" : "22177878", "abstract" : "Cancer control and toxicity outcomes are the mainstay of evidence-based medicine in radiation oncology. However, radiotherapy is an intricate therapy involving numerous processes that need to be executed appropriately in order for the therapy to be delivered successfully. The use of image-guided radiation therapy (IGRT), referring to imaging occurring in the radiation therapy room with per-patient adjustments, can increase the agreement between the planned and the actual dose delivered. However, the absence of direct evidence regarding the clinical benefit of IGRT has been a criticism. Here, we dissect the role of IGRT in the radiotherapy (RT) process and emphasize its role in improving the quality of the intervention. The literature is reviewed to collect evidence that supports that higher-quality dose delivery enabled by IGRT results in higher clinical control rates, reduced toxicity, and new treatment options for patients that previously were without viable options. \u00a9 2012.", "author" : [ { "dropping-particle" : "", "family" : "Bujold", "given" : "Alexis", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Craig", "given" : "Tim", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Jaffray", "given" : "David", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Dawson", "given" : "Laura A.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Seminars in Radiation Oncology", "id" : "ITEM-1", "issue" : "1", "issued" : { "date-parts" : [ [ "2012" ] ] }, "page" : "50-61", "publisher" : "Elsevier Inc.", "title" : "Image-Guided Radiotherapy: Has It Influenced Patient Outcomes?", "type" : "article-journal", "volume" : "22" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[14]", "plainTextFormattedCitation" : "[14]", "previouslyFormattedCitation" : "[14]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[14] and Jaffray et al ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1038/nrclinonc.2012.194", "ISBN" : "1759-4774\\r1759-4782", "ISSN" : "1759-4774", "PMID" : "23165124", "abstract" : "Radiotherapy is a highly effective, targeted therapy for the management of cancer. Technological innovations have enabled the direct integration of imaging technology into the radiation treatment device to increase the precision and accuracy of radiation delivery. As well as addressing a clinical need to better control the placement of the dose within the body, image-guided radiotherapy has enabled innovators in the field to accelerate their exploration of a number of different paradigms of radiation delivery, including toxicity reduction, dose escalation, hypofractionation, voxelization, and adaptation. Although these approaches are already innovative trends in radiation oncology, it is anticipated that they will work synergistically with other innovations in cancer management (including biomarker strategies, novel systemic and local therapies) as part of the broader goal of personalized cancer medicine. This Review discusses the rationale for adopting image-guidance approaches in radiotherapy, and the technology for achieving precision and accuracy in the context of different paradigms within the evolving radiation oncology practice. It also examines exciting advances in radiotherapy technology that suggest a convergence of radiotherapy practice in which patient-specific radiotherapy treatment courses are one of the most personalized forms of intervention in cancer medicine.", "author" : [ { "dropping-particle" : "", "family" : "Jaffray", "given" : "David A.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Nature Reviews Clinical Oncology", "id" : "ITEM-1", "issue" : "12", "issued" : { "date-parts" : [ [ "2012" ] ] }, "page" : "688-699", "publisher" : "Nature Publishing Group", "title" : "Image-guided radiotherapy: from current concept to future perspectives", "type" : "article-journal", "volume" : "9" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[15]", "plainTextFormattedCitation" : "[15]", "previouslyFormattedCitation" : "[15]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[15], due to the assumed benefit of IGRT, quantifying its efficacy in prospective clinical trials is problematic given the difficulty in arguing equipoise between the two arms. In this context retrospective observational studies may provide insight.Following its immediate clinical use, IGRT data is generally archived in accordance with national data retention guidelines, and is rarely analysed further. In this paper we hypothesize that IGRT data have the potential to provide much deeper insight. While some simulation studies have been performed to look at the effect of residual setup errors from different imaging protocols, such as the work by Han et al ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/j.radonc.2011.07.027", "ISSN" : "1879-0887", "PMID" : "21872956", "abstract" : "BACKGROUND AND PURPOSE: To evaluate residual patient setup errors and daily dose variations of different less-than-daily image guidance (IG) strategies in the delivery of external beam radiotherapy for esophageal cancer.\n\nMATERIAL AND METHODS: Daily image-guided setup data for 25 consecutive esophageal cancer patients treated with helical tomotherapy were evaluated. Seven less-than-daily IG strategies with different imaging frequencies were simulated. For each IG strategy, the daily residual setup errors were calculated. Using TomoTherapy Planned Adaptive software, daily dose variations to the clinical target volume, heart, and lungs were evaluated in five representative patients.\n\nRESULTS: With 0% (60%) IG frequency, the margins required for adequate coverage of the clinical target volume were 13 mm (10 mm), 14 mm (11 mm), and 5 mm (5 mm) in the left-right, superior-inferior, and anterior-posterior directions, respectively. Even with 60% IG frequency, 10% of the fractions had more than 10% decrease in the dose level covering 95% of the target, and 14% and 13% of the fractions had more than 10% increase in total lung volume receiving at least 0.8 Gy per fraction, and heart volume receiving at least 1.2 Gy per fraction, respectively.\n\nCONCLUSION: Substantial residual setup errors would occur for treatment fractions without IG even if the most frequent less-than-daily IG strategy was to be used, which could lead to significant daily dose variations for the target volume and adjacent normal tissues. Daily image guidance is recommended throughout the course of treatment in conformal radiotherapy for esophageal cancer.", "author" : [ { "dropping-particle" : "", "family" : "Han", "given" : "Chunhui", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Schiffner", "given" : "Daniel C", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Schultheiss", "given" : "Timothy E", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Chen", "given" : "Yi-Jen", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Liu", "given" : "An", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Wong", "given" : "Jeffrey Y C", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Radiotherapy and oncology : journal of the European Society for Therapeutic Radiology and Oncology", "id" : "ITEM-1", "issue" : "2", "issued" : { "date-parts" : [ [ "2012", "2" ] ] }, "page" : "309-14", "publisher" : "Elsevier Ireland Ltd", "title" : "Residual setup errors and dose variations with less-than-daily image guided patient setup in external beam radiotherapy for esophageal cancer.", "type" : "article-journal", "volume" : "102" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[8]", "plainTextFormattedCitation" : "[8]", "previouslyFormattedCitation" : "[8]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[8], to the best of our knowledge, no studies have yet directly explored the relationship between setup error data and patient outcomes. We analysed a cohort of 780 Non-Small Cell Lung Cancer (NSCLC) patients who were treated with curative intent radiotherapy and verified by CBCT based IGRT. The aim was to assess: i) whether the magnitude of residual set-up errors following IGRT is associated with patient survival, and ii) should any such relationship exist, if the directionality of the residual errors can provide evidence of the underlying cause.Materials and MethodsAnonymised routine clinical data for 780 NSCLC patients and 167 oesophagus patients, treated with curative intent, were collated with the approval of the UK Health Research Authority and the local Caldicott information governance committee (ethical approval ref. 17/NW/0060). Lung patients were treated with standard fractionation regimes of either 55Gy in 20 fractions or 60 – 66Gy in 30 – 33 fractions, with or without chemotherapy delivered sequentially or concurrently. Oesophageal patients received either 50Gy in 25 fractions, 55Gy in 20 fractions or 41.4Gy in 23 fractions, with or without concurrent chemotherapy. The IGRT protocol was identical for both patient cohorts: CBCT images were acquired at the first 3 fractions and weekly thereafter, and registered to the planning scan using a rigid registration based on bony-anatomy, by placing a region of interest over the spine (XVI version 4.0 or 5.0). The obtained translations and rotations from the image match were then applied to the centre of the PTV to derive the appropriate couch shift. If any of the required shifts were greater than the 5mm action threshold in any direction, then an online correction was performed. A systematic correction to the patient setup for subsequent fractions was made after the first 3 fractions if the average shift in the first three fractions exceeded the 5mm action level. If a systematic correction was applied then further images were acquired to verify its execution. Imaging was then subsequently performed weekly to verify the patient position with respect to possible positional time-trends, and online corrections and further monitoring was performed if required.The image registration translations were collected for each patient along with routine clinical variables, including: patient age, gender, performance status, comorbidities score, stage, GTV volume and the fractionation regime, as detailed in Table 1. The GTV variable consisted of both true GTVs, and ITVs eroded using a kernel developed in a dataset with both volumes available, as described by Johnson et al ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/j.radonc.2017.11.015", "ISSN" : "18790887", "abstract" : "Background and purpose: The gross tumour volume (GTV) is predictive of clinical outcome and consequently features in many machine-learned models. 4D-planning, however, has prompted substitution of the GTV with the internal gross target volume (iGTV). We present and validate a method to synthesise GTV data from the iGTV, allowing the combination of 3D and 4D planned patient cohorts for modelling. Material and methods: Expert delineations in 40 non-small cell lung cancer patients were used to develop linear fit and erosion methods to synthesise the GTV volume and shape. Quality was assessed using Dice Similarity Coefficients (DSC) and closest point measurements; by calculating dosimetric features; and by assessing the quality of random forest models built on patient populations with and without synthetic GTVs. Results: Volume estimates were within the magnitudes of inter-observer delineation variability. Shape comparisons produced mean DSCs of 0.8817 and 0.8584 for upper and lower lobe cases, respectively. A model trained on combined true and synthetic data performed significantly better than models trained on GTV alone, or combined GTV and iGTV data. Conclusions: Accurate synthesis of GTV size from the iGTV permits the combination of lung cancer patient cohorts, facilitating machine learning applications in thoracic radiotherapy.", "author" : [ { "dropping-particle" : "", "family" : "Johnson", "given" : "Corinne", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Price", "given" : "Gareth", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Khalifa", "given" : "Jonathan", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Faivre-Finn", "given" : "Corinne", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Dekker", "given" : "Andre", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Moore", "given" : "Christopher", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Herk", "given" : "Marcel", "non-dropping-particle" : "van", "parse-names" : false, "suffix" : "" } ], "container-title" : "Radiotherapy and Oncology", "id" : "ITEM-1", "issue" : "2", "issued" : { "date-parts" : [ [ "2018" ] ] }, "page" : "355-361", "publisher" : "Elsevier B.V.", "title" : "A method to combine target volume data from 3D and 4D planned thoracic radiotherapy patient cohorts for machine learning applications", "type" : "article-journal", "volume" : "126" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[16]", "plainTextFormattedCitation" : "[16]", "previouslyFormattedCitation" : "[16]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[16]. The natural logarithm transform of the GTV was taken to normalise the data. Variables with a large percentage of missing data (> 15% missing) were excluded and remaining missing data was imputed using a random forest method. Each patient’s shifts were processed as shown in Figure 1a. First, the 5mm action threshold was applied to the raw shift data to give a measure of the residual setup error. Couch corrections were assumed to be performed perfectly, due to the lack of data on the couch start and end positions. Any systematic corrections applied to the patient setup were ignored, as these changes will be reflected in shift values obtained in later treatment fractions. A nearest-neighbour weighting approach was then used to backfill fractions where no image was acquired in order to remove any bias resulting from overrepresentation of the first 3 fractions (which are always imaged for every patient) and incorporate any longitudinal trends in the errors ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/J.IJROBP.2006.11.050", "ISSN" : "0360-3016", "author" : [ { "dropping-particle" : "", "family" : "Boer", "given" : "Hans C J", "non-dropping-particle" : "de", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Heijmen", "given" : "Ben J M", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "International Journal of Radiation Oncology*Biology*Physics", "id" : "ITEM-1", "issue" : "5", "issued" : { "date-parts" : [ [ "2007", "4", "1" ] ] }, "page" : "1586-1595", "publisher" : "Elsevier", "title" : "eNAL: An Extension of the NAL Setup Correction Protocol for Effective Use of Weekly Follow-up Measurements", "type" : "article-journal", "volume" : "67" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[10]", "plainTextFormattedCitation" : "[10]", "previouslyFormattedCitation" : "[10]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[10]. The weighted residual shift data were then summarised as a number of ‘shift metrics’, including: the mean and standard deviation of the shifts in the lateral, longitudinal and vertical directions; and the vector shift to the heart, represented schematically in Figure 1b, using the centre of mass (CoM) as a representative point for the heart position. As heart delineations are not available for all patients, we used a linear fit to synthesise the heart CoM from the lung CoM (as lung contours were available for all patients). Briefly, using a cohort of 554 lung patients, all of whom had both a lung and a heart contour available, plots of the lung versus heart centre of mass position in each of the x, y and z directions were created and linear fits determined. These linear fits were then used to convert lung CoM coordinates to heart CoM coordinates. The linear fit quality was assessed via the coefficient of determination (R2) and standard deviation of the residuals. Further details are provided in the supplementary data. All data was analysed in International Electrotechnical Commission (IEC) coordinates, such that positive shifts in the x, y and z directions corresponds to shifts to the patient left, superior and anterior, respectively.The main analysis took place in the NSCLC cohort. Univariate analysis of any relationship between clinical variables and the shift metrics was undertaken via Pearson correlation and Analysis of Variance. The resulting p-values were adjusted using the Benjamini and Hochberg False Discovery Rate (FDR) method ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "author" : [ { "dropping-particle" : "", "family" : "Benjamini", "given" : "Yoav", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Hochberg", "given" : "Yosef", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Journal of the Royal Statistical Society: Series B", "id" : "ITEM-1", "issue" : "1", "issued" : { "date-parts" : [ [ "1995" ] ] }, "page" : "289-300", "title" : "Controlling the False Discovery Rate : A Practical and Powerful Approach to Multiple Testing", "type" : "article-journal", "volume" : "57" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[17]", "plainTextFormattedCitation" : "[17]", "previouslyFormattedCitation" : "[17]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[17], to correct for the effects of multiple comparisons. A separate comparison was performed for the comorbidities score, using the subset of the data with a comorbidity score available (n = 345), owing to the large amount of missing data. Elastic net penalized Cox regression with equal ridge regression and LASSO penalty terms was used to select the variables most strongly related to patients’ overall survival. The most significant residual shift variables were used to categorise patients into low and high shift groups, using an optimal cut-point determined by maximising the log-rank statistic between the two groups. Kaplan-Meier survival curves for each group were plotted and multivariate analysis of both the categorized risk groups and the continuous residual shift magnitudes alongside well known prognostic factors was performed using Cox regression. Initial analysis looked at the whole cohort, and then at subsets grouped by tumour position relative to the heart CoM to determine any directional effect. These subsets were limited to patients for whom both a GTV/ITV and heart/lung delineation were available (n=628). Additionally, the vector shift to the heart (towards or away) was evaluated. Results were validated in the oesophagus cancer cohort using the same methodology. All analysis was performed in R version 3.3.2 ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "author" : [ { "dropping-particle" : "", "family" : "R Core Team", "given" : "", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "id" : "ITEM-1", "issued" : { "date-parts" : [ [ "2016" ] ] }, "publisher" : "R Foundation for Statistical Computing", "publisher-place" : "Vienna, Austria", "title" : "R: A Language and Environment for Statistical Computing", "type" : "article" }, "uris" : [ "" ] }, { "id" : "ITEM-2", "itemData" : { "author" : [ { "dropping-particle" : "", "family" : "Ishwaran", "given" : "Hemant", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Kogalur", "given" : "Udaya B", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "id" : "ITEM-2", "issued" : { "date-parts" : [ [ "2016" ] ] }, "number" : "R package version 2.2.0.", "title" : "Random Forests for Survival, Regression and Classification (RF-SRC)", "type" : "article" }, "uris" : [ "" ] }, { "id" : "ITEM-3", "itemData" : { "author" : [ { "dropping-particle" : "", "family" : "Ishwaran", "given" : "Hemant", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Kogalur", "given" : "UB", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "RNews", "id" : "ITEM-3", "issue" : "2", "issued" : { "date-parts" : [ [ "2007" ] ] }, "page" : "25-31", "title" : "Random survival forests for R", "type" : "article-journal", "volume" : "7" }, "uris" : [ "" ] }, { "id" : "ITEM-4", "itemData" : { "DOI" : "10.1214/08-AOAS169", "ISSN" : "1932-6157", "author" : [ { "dropping-particle" : "", "family" : "Ishwaran", "given" : "Hemant", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Kogalur", "given" : "Udaya B.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Blackstone", "given" : "Eugene H.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Lauer", "given" : "Michael S.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "The Annals of Applied Statistics", "id" : "ITEM-4", "issue" : "3", "issued" : { "date-parts" : [ [ "2008", "9" ] ] }, "page" : "841-860", "title" : "Random survival forests", "type" : "article-journal", "volume" : "2" }, "uris" : [ "" ] }, { "id" : "ITEM-5", "itemData" : { "author" : [ { "dropping-particle" : "", "family" : "Kassambara", "given" : "Alboukadel", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "id" : "ITEM-5", "issued" : { "date-parts" : [ [ "2016" ] ] }, "number" : "0.2.4", "title" : "Survminer: Drawing Survival Curves using 'ggplot2'", "type" : "article" }, "uris" : [ "" ] }, { "id" : "ITEM-6", "itemData" : { "ISBN" : "0-387-98784-3", "author" : [ { "dropping-particle" : "", "family" : "Therneau", "given" : "Terry M.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Grambsch", "given" : "Patricia M.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "id" : "ITEM-6", "issued" : { "date-parts" : [ [ "2000" ] ] }, "publisher" : "Springer", "publisher-place" : "New York", "title" : "Modeling Survival Data: Extending the Cox Model", "type" : "book" }, "uris" : [ "" ] }, { "id" : "ITEM-7", "itemData" : { "author" : [ { "dropping-particle" : "", "family" : "Therneau", "given" : "Terry M.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Lumley", "given" : "Thomas", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "id" : "ITEM-7", "issued" : { "date-parts" : [ [ "2015" ] ] }, "number" : "2.38", "title" : "A Package for Survival Analysis in S", "type" : "article" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[18\u201324]", "plainTextFormattedCitation" : "[18\u201324]", "previouslyFormattedCitation" : "[18\u201324]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[18–24].ResultsThe volume of missing data for each variable is reported in Table 1. All variables had an acceptable level of missing data apart from the comorbidities score which was excluded from subsequent analysis. NSCLC DataUnivariate analysis found the shift variables to be independent of the clinical variables listed in Table 1. Additionally, no relation was observed between shift and comorbidities in the subset of patients for whom a comorbidity score was available (n=345). Representative plots along with their corresponding Analysis of Variance p-values and Pearson correlation coefficient (where relevant) are shown in Figure 2.Variable selection found the standard deviation of the lateral shift, the vector shift to the heart, age, ECOG performance status, the fractionation and the GTV volume to be significantly related to overall survival. When the shift direction towards or away from the heart was taken into account, the mean longitudinal, vertical and lateral shifts were also selected in addition to those listed previously. Table 2 details the split points used to categorise the risk and the multivariate analysis results for the different shift metrics selected as significant by the regularised Cox regression. Results from the inferior and anterior cohort results are not provided due to the limited cohort sizes (n = 62 and 29, respectively). It is clear from Table 2 that the right and left cohorts show opposite effects. In the left sided tumour cohort, cases with mean residual shifts greater than 0.1mm decreased overall survival. While in the right sided tumour cohort, cases with mean residual shifts greater than 0.4mm have improved overall survival. This result (visualised in Figure 3a and b) suggests there is a survival effect depending on whether the dose is being shifted towards or away from the mediastinum. Similar effects are seen in the superior and posterior cohorts, with cases that have mean shifts towards the mediastinum having worse overall survival. The effect of the magnitude of the three-dimensional residual shift, evaluated as the vector shift to the heart (range -4.34mm – 4.66mm, mean -0.09mm), on overall survival is shown in Table 2 and Figure 3c, where it can be seen that the patients with shifts towards the heart CoM have significantly worse overall survival. Studying the vector shift to the heart as a continuous variable, we find a multivariate Cox corrected p-value of 0.007 and hazard ratio (HR) of 1.091 per mm shift, demonstrating an increased risk of death with increasing shifts towards the heart. The same analysis using the overall magnitude of the mean shift, i.e. the undirected residual set-up errors, yields a HR of 0.998 (p=0.737), demonstrating that the magnitude alone is not correlated with patient survival.Oesophagus DataThe vector shift to the heart in the oesophagus cancer cohort was found to range between 3.85mm and 4.16mm (mean -0.13mm). Using a split point of 0.0mm on the vector shift to the heart (as for the lung cohort) resulted in 82 patients falling into the high shift category, and 95 falling into the low shift category. The survival curves for the high and low shift groups are shown in Figure 4.As for the NSCLC cases, it can be seen that the patients who have the higher vector shift to the heart values have worse overall survival. This result is statistically significant when corrected for potential confounding factors (HR (low shift) = 0.660, p = 0.029). When the vector shift to the heart is used as a continuous variable, the hazard ratio is 1.164 per mm, p = 0.041.DiscussionThis study explored whether image-guided radiotherapy residual setup errors directly correlate with overall patient survival after curative intent radiotherapy. In a cohort of 780 NSCLC patients, whilst it was found that the magnitude of the residual shifts over the course of treatment was not associated with poorer survival, the overall shift towards or away from the heart had a significant effect on overall patient survival (Figure 3). The hazard ratio found when studying the vector shift to the heart as a continuous variable was 1.091 per mm, demonstrating that greater shifts towards the heart result in a greater risk (where positive shifts equate to the heart being moved towards the high dose region, and negative shifts moving the heart away). The same relation was observed when the shifts were normalised to the proximity of the tumour to the heart (see supplementary material), and additionally in an oesophageal cancer patient cohort (n = 177), using the same split point as the lung cohort, suggesting that the effect of small residual errors is not limited to one particular cancer subtype. Univariate analysis showed that no patient variables were significantly related to any of the selected shift metrics, suggesting that the residual shift acts as an independent factor. All survival results were still significant when corrected for the effects of well-known prognostic factors.As far as the authors are aware, this is the first study to quantify the effect of residual errors on patient outcomes. A previous study by Han et al ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/j.radonc.2011.07.027", "ISSN" : "1879-0887", "PMID" : "21872956", "abstract" : "BACKGROUND AND PURPOSE: To evaluate residual patient setup errors and daily dose variations of different less-than-daily image guidance (IG) strategies in the delivery of external beam radiotherapy for esophageal cancer.\n\nMATERIAL AND METHODS: Daily image-guided setup data for 25 consecutive esophageal cancer patients treated with helical tomotherapy were evaluated. Seven less-than-daily IG strategies with different imaging frequencies were simulated. For each IG strategy, the daily residual setup errors were calculated. Using TomoTherapy Planned Adaptive software, daily dose variations to the clinical target volume, heart, and lungs were evaluated in five representative patients.\n\nRESULTS: With 0% (60%) IG frequency, the margins required for adequate coverage of the clinical target volume were 13 mm (10 mm), 14 mm (11 mm), and 5 mm (5 mm) in the left-right, superior-inferior, and anterior-posterior directions, respectively. Even with 60% IG frequency, 10% of the fractions had more than 10% decrease in the dose level covering 95% of the target, and 14% and 13% of the fractions had more than 10% increase in total lung volume receiving at least 0.8 Gy per fraction, and heart volume receiving at least 1.2 Gy per fraction, respectively.\n\nCONCLUSION: Substantial residual setup errors would occur for treatment fractions without IG even if the most frequent less-than-daily IG strategy was to be used, which could lead to significant daily dose variations for the target volume and adjacent normal tissues. Daily image guidance is recommended throughout the course of treatment in conformal radiotherapy for esophageal cancer.", "author" : [ { "dropping-particle" : "", "family" : "Han", "given" : "Chunhui", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Schiffner", "given" : "Daniel C", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Schultheiss", "given" : "Timothy E", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Chen", "given" : "Yi-Jen", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Liu", "given" : "An", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Wong", "given" : "Jeffrey Y C", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Radiotherapy and oncology : journal of the European Society for Therapeutic Radiology and Oncology", "id" : "ITEM-1", "issue" : "2", "issued" : { "date-parts" : [ [ "2012", "2" ] ] }, "page" : "309-14", "publisher" : "Elsevier Ireland Ltd", "title" : "Residual setup errors and dose variations with less-than-daily image guided patient setup in external beam radiotherapy for esophageal cancer.", "type" : "article-journal", "volume" : "102" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[8]", "plainTextFormattedCitation" : "[8]", "previouslyFormattedCitation" : "[8]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[8] looked at quantifying the dose variations in oesophageal cancer caused by residual setup errors when different less-than-daily image guidance protocols were used. They found that even when the most frequent less-than-daily protocol was used, residual errors resulted in an under-dosing of the CTV and over-dosing of the heart and lungs. They did not however, quantify the effect of these dose differences on patient survival. Recent studies have discussed the detrimental impact of heart dose on survival. Notably, the RTOG 0617 phase 3 study ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/S1470-2045(14)71207-0", "ISSN" : "14702045", "author" : [ { "dropping-particle" : "", "family" : "Bradley", "given" : "Jeffrey D", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Paulus", "given" : "Rebecca", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Komaki", "given" : "Ritsuko", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Masters", "given" : "Gregory", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Blumenschein", "given" : "George", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Schild", "given" : "Steven", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Bogart", "given" : "Jeffrey", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Hu", "given" : "Chen", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Forster", "given" : "Kenneth", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Magliocco", "given" : "Anthony", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Kavadi", "given" : "Vivek", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Garces", "given" : "Yolanda I", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Narayan", "given" : "Samir", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Iyengar", "given" : "Puneeth", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Robinson", "given" : "Cliff", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Wynn", "given" : "Raymond B", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Koprowski", "given" : "Christopher", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Meng", "given" : "Joanne", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Beitler", "given" : "Jonathan", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Gaur", "given" : "Rakesh", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Curran", "given" : "Walter", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Choy", "given" : "Hak", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "The Lancet Oncology", "id" : "ITEM-1", "issue" : "Rtog 0617", "issued" : { "date-parts" : [ [ "2015", "1" ] ] }, "page" : "187-199", "title" : "Standard-dose versus high-dose conformal radiotherapy with concurrent and consolidation carboplatin plus paclitaxel with or without cetuximab for patients with stage IIIA or IIIB non-small-cell lung cancer (RTOG 0617): a randomised, two-by-two factorial p", "type" : "article-journal" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[25]", "plainTextFormattedCitation" : "[25]", "previouslyFormattedCitation" : "[25]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[25], which compared standard- and high-dose radiotherapy in stage III NSCLC patients, reported worse survival in the high-dose arm. This result seemed counter-intuitive as a higher dose was expected to improve local control. However, higher lung and heart doses were also reported in the high dose arm, which has been suggested to be one possible cause of this survival difference ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/S1470-2045(15)70001-X", "ISSN" : "1474-5488", "PMID" : "25601340", "author" : [ { "dropping-particle" : "", "family" : "Faivre-Finn", "given" : "Corinne", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "The Lancet. Oncology", "id" : "ITEM-1", "issue" : "2", "issued" : { "date-parts" : [ [ "2015", "2" ] ] }, "page" : "125-7", "title" : "Dose escalation in lung cancer: have we gone full circle?", "type" : "article-journal", "volume" : "16" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[26]", "plainTextFormattedCitation" : "[26]", "previouslyFormattedCitation" : "[26]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[26]. A recent study performed by McWilliam et al. used image-based data mining methods to identify anatomical regions where the radiation dose correlated with survival ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/j.ejca.2017.07.053", "ISSN" : "0959-8049", "author" : [ { "dropping-particle" : "", "family" : "Mcwilliam", "given" : "Alan", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Kennedy", "given" : "Jason", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Hodgson", "given" : "Clare", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Vasquez", "given" : "Eliana", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Faivre-finn", "given" : "Corinne", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Herk", "given" : "Marcel", "non-dropping-particle" : "van", "parse-names" : false, "suffix" : "" } ], "container-title" : "European Journal of Cancer", "id" : "ITEM-1", "issued" : { "date-parts" : [ [ "2017" ] ] }, "page" : "106-113", "publisher" : "Elsevier Ltd", "title" : "Radiation dose to heart base linked with poorer survival in lung cancer patients", "type" : "article-journal", "volume" : "85" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[27]", "plainTextFormattedCitation" : "[27]", "previouslyFormattedCitation" : "[27]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[27]. The most significant difference was found in the base of the heart, and when the mean dose to the heart was used to split the patients into high and low dose groups, a significant difference in survival was found.Our study has several limitations. Firstly, the number of CBCTs per patient varied. To correct for this fact, we employed a nearest-neighbour weighting approach. Whilst this is a relatively robust method of compensating for the over-representation of the first three fractions, it will not perfectly reflect the variations seen in patient position on a day to day basis. The effects of such imperfections however, should be matched between the two groups we compare. Nonetheless, validation in a cohort with daily imaging is desirable to confirm this method. Additionally, the residual setup errors are currently obtained by applying the imaging action threshold retrospectively to the recorded shift data, which assumes that the couch correction is always perfect. It would be preferable if patient couch positions were recorded at each fraction, but this was not the case. Over this large cohort of patients we assume that the effect of imperfect corrections will be equivalent in both groups, as the cohort is split at a 0mm shift. This method of estimating the residual shifts is likely to underestimate the residuals to a certain extent ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/j.ijrobp.2008.08.059", "ISBN" : "0360-3016", "ISSN" : "03603016", "PMID" : "19095368", "abstract" : "Purpose To determine the geometric accuracy of lung cancer radiotherapy using daily volumetric, cone-beam CT (CBCT) image guidance and online couch position adjustment. Methods and Materials Initial setup accuracy using localization CBCT was analyzed in three lung cancer patient cohorts. The first (n = 19) involved patients with early-stage non\u2013small-cell lung cancer (NSCLC) treated using stereotactic body radiotherapy (SBRT). The second (n = 48) and third groups (n = 20) involved patients with locally advanced NSCLC adjusted with manual and remote-controlled couch adjustment, respectively. For each group, the couch position was adjusted when positional discrepancies exceeded \u00b13 mm in any direction, with the remote-controlled couch correcting all three directions simultaneously. Adjustment accuracy was verified with a second CBCT. Population-based setup margins were derived from systematic (\u03a3) and random (\u03c3) positional errors for each group. Results Localization imaging demonstrates that 3D positioning errors exceeding 5 mm occur in 54.5% of all delivered fractions. CBCT reduces these errors; post-correction \u03a3 and \u03c3 ranged from 1.2 to 1.9 mm for Group 1, with 82% of all fractions within \u00b13 mm. For Group 2, \u03a3 and \u03c3 ranged between 0.8 and 1.8 mm, with 76% of all treatment fractions within \u00b13 mm. For Group 3, the remote-controlled couch raised this to 84%, and \u03a3 and \u03c3 were reduced to 0.4 to 1.7 mm. For each group, the postcorrection setup margins were 4 to 6 mm, 3 to 4 mm, and 2 to 3 mm, respectively. Conclusions Using IGRT, high geometric accuracy is achievable for NSCLC patients, potentially leading to reduced PTV margins, improved outcomes and empowering adaptive radiation therapy for lung cancer.", "author" : [ { "dropping-particle" : "", "family" : "Bissonnette", "given" : "Jean-Pierre", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Purdie", "given" : "Thomas G.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Higgins", "given" : "Jane A.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Li", "given" : "Winnie", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Bezjak", "given" : "Andrea", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "International Journal of Radiation Oncology*Biology*Physics", "id" : "ITEM-1", "issue" : "3", "issued" : { "date-parts" : [ [ "2009" ] ] }, "page" : "927-934", "title" : "Cone-Beam Computed Tomographic Image Guidance for Lung Cancer Radiation Therapy", "type" : "article-journal", "volume" : "73" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[6]", "plainTextFormattedCitation" : "[6]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[6]. However, this underestimation is unbiased and will not therefore affect the primary finding i.e. that shifts towards and away from the heart affect survival. However, this bias will affect the results presented in Table 2, so they must be interpreted with care. Secondly, our survival analysis has not been being corrected for comorbidities. However, in the cohort where comorbidity scores were available, no significant relation was found between comorbidity score and any shift variables. Analysis should be re-performed in a larger cohort to confirm that no relation exists, and to allow the comorbidity score to be included in the multivariate correction procedure in case of any unexpected variable interactions.Thirdly, the centre of mass of the heart was chosen as a representative point for categorisation. It could be that other reference points are more informative, such as the base of the heart, as discussed by McWilliam et al. ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.1016/j.ejca.2017.07.053", "ISSN" : "0959-8049", "author" : [ { "dropping-particle" : "", "family" : "Mcwilliam", "given" : "Alan", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Kennedy", "given" : "Jason", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Hodgson", "given" : "Clare", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Vasquez", "given" : "Eliana", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Faivre-finn", "given" : "Corinne", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Herk", "given" : "Marcel", "non-dropping-particle" : "van", "parse-names" : false, "suffix" : "" } ], "container-title" : "European Journal of Cancer", "id" : "ITEM-1", "issued" : { "date-parts" : [ [ "2017" ] ] }, "page" : "106-113", "publisher" : "Elsevier Ltd", "title" : "Radiation dose to heart base linked with poorer survival in lung cancer patients", "type" : "article-journal", "volume" : "85" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[27]", "plainTextFormattedCitation" : "[27]", "previouslyFormattedCitation" : "[27]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[27]. This will be explored, as part of a full dosimetric analysis.Lastly, besides the primary endpoint used in this study of overall survival, it would have been of interest to investigate cardiac toxicity as an endpoint. Unfortunately this was not possible as the prospective recording of toxicities (including cardiac) in the routine setting is not robust enough. A particular issue is the recording of treatment-related cardiac deaths which tend to be attributed to lung cancer by general practitioners, without further investigations.Other factors that could affect patient survival, such as the level of tumour shrinkage that occurs throughout the course of treatment, should also be investigated. However, due to the independence of the residual shift from all of the clinical parameters we tested, we argue that it is fair to assume the distribution of tumour shrinkage and other effects to be comparable in both patient groups, and hence that our results should not be unduly biased by any such effects.As discussed in the introduction, there is little previous evidence linking IGRT directly with patient outcomes, which is largely due to the assumed benefit and therefore difficulty in arguing equipoise for a clinical trial testing its benefit. This retrospective study provides some of the first direct evidence of the clinical impact of IGRT, showing that even small residual set-up errors impact lung cancer patient survival. Based on our results, we would suggest firstly that a 5mm action threshold is too high, as the patients who have deviations that are not large enough to be corrected have the worst survival. Our results suggest that this effect is due to toxicity and not tumour control, as the magnitude of the residual errors alone does not correlate with survival. Secondly, we would advise that further investigations into the effect of dose to sub-structures of the heart are conducted and that stricter constraints are put in place. Currently planning constraints for the heart are used (V30Gy < 40% and V40Gy < 30%), but these are compromised in favour of target coverage. Heart sparing strategies such as deep inspiration breath-hold are being investigated for lung cancer ADDIN CSL_CITATION { "citationItems" : [ { "id" : "ITEM-1", "itemData" : { "DOI" : "10.3109/0284186X.2016.1142115", "ISBN" : "0284-186X", "ISSN" : "1651226X", "PMID" : "26935017", "author" : [ { "dropping-particle" : "", "family" : "Persson", "given" : "Gitte F.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Scherman Rydh\u00f6g", "given" : "Jonas", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Josipovic", "given" : "Mirjana", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "V.", "family" : "Maraldo", "given" : "Maja", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Nyg\u00e5rd", "given" : "Lotte", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Costa", "given" : "Junia", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Berthelsen", "given" : "Anne K.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Specht", "given" : "Lena", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" }, { "dropping-particle" : "", "family" : "Aznar", "given" : "Marianne C.", "non-dropping-particle" : "", "parse-names" : false, "suffix" : "" } ], "container-title" : "Acta Oncologica", "id" : "ITEM-1", "issue" : "8", "issued" : { "date-parts" : [ [ "2016" ] ] }, "page" : "1053-1056", "title" : "Deep inspiration breath-hold volumetric modulated arc radiotherapy decreases dose to mediastinal structures in locally advanced lung cancer", "type" : "article-journal", "volume" : "55" }, "uris" : [ "" ] } ], "mendeley" : { "formattedCitation" : "[28]", "plainTextFormattedCitation" : "[28]", "previouslyFormattedCitation" : "[28]" }, "properties" : { "noteIndex" : 0 }, "schema" : "" }[28] however, their effectiveness seems to be highly patient dependant. In the last year the standard treatment imaging protocol at our institution has been updated, such that now, CBCT images are acquired at every fraction, and the action level has been reduced to 2mm. We intend to repeat this analysis in this new cohort, once sufficient outcome data has been collected, to confirm whether or not this change has removed the observed survival effect. The origin of the survival difference seems most likely to be linked to increasing/decreasing dose to the heart. We are now investigating the effect of residual setup errors on the cumulative dose.ConclusionIn this study we have shown that residual IGRT shifts significantly correlate with survival. It was found that patients who have a mean residual shift towards the heart have a worse prognosis as compared to those who have a mean shift away from the heart. This effect was observed in both non-small cell lung cancer and oesophageal cancer cohorts. These results provide solid evidence for the use of stricter IGRT protocols for thoracic radiotherapy, as they show that even small residuals have a significant effect on survival, and provide further information on the dose response of OARs. We recommend that imaging action thresholds are reviewed, along with radiotherapy heart constraints, as increasing dose to the heart appears to have an early effect on survival. ReferencesADDIN Mendeley Bibliography CSL_BIBLIOGRAPHY [1]Cancer Research UK. Cancer Statistics - Worldwide Cancer n.d. (accessed November 6, 2017).[2]Cancer Research UK. About Cancer - Lung Cancer n.d. (accessed March 23, 2017).[3]Peters LJ, O’Sullivan B, Giralt J, Fitzgerald TJ, Trotti A, Bernier J, et al. Critical impact of radiotherapy protocol compliance and quality in the treatment of advanced head and neck cancer: Results from TROG 02.02. J Clin Oncol 2010;28:2996–3001. doi:10.1200/JCO.2009.27.4498.[4]De Crevoisier R, Tucker SL, Dong L, Mohan R, Cheung R, Cox JD, et al. Increased risk of biochemical and local failure in patients with distended rectum on the planning CT for prostate cancer radiotherapy. Int J Radiat Oncol Biol Phys 2005;62:965–73. doi:10.1016/j.ijrobp.2004.11.032.[5]Dawson LA, Sharpe MB. Image-guided radiotherapy?: rationale , benefits , and limitations. Lancet Oncol 2006;7:848–58. doi:10.1016/S1470-2045(06)70904-4.[6]Bissonnette J-P, Purdie TG, Higgins JA, Li W, Bezjak A. Cone-Beam Computed Tomographic Image Guidance for Lung Cancer Radiation Therapy. Int J Radiat Oncol 2009;73:927–34. doi:10.1016/j.ijrobp.2008.08.059.[7]Li W, Moseley DJ, Bissonnette J-P, Purdie TG, Bezjak A, Jaffray DA. Setup Reproducibility for Thoracic and Upper Gastrointestinal Radiation Therapy: Influence of Immobilization Method and On-Line Cone-Beam CT Guidance. Med Dosim 2010;35:287–96. doi:10.1016/j.meddos.2009.09.003.[8]Han C, Schiffner DC, Schultheiss TE, Chen Y-J, Liu A, Wong JYC. Residual setup errors and dose variations with less-than-daily image guided patient setup in external beam radiotherapy for esophageal cancer. Radiother Oncol 2012;102:309–14. doi:10.1016/j.radonc.2011.07.027.[9]de Boer HCJ, Heijmen BJM. A Protocol for the Reduction of Systematic Patient Setup Errors with Minimal Portal Imaging Workload. Int J Radiat Oncol 2001;50:1350–65. doi:10.1016/S0360-3016(01)01624-8.[10]de Boer HCJ, Heijmen BJM. eNAL: An Extension of the NAL Setup Correction Protocol for Effective Use of Weekly Follow-up Measurements. Int J Radiat Oncol 2007;67:1586–95. doi:10.1016/J.IJROBP.2006.11.050.[11]Nijkamp J, Pos FJ, Nuver TT, de Jong R, Remeijer P, Sonke J-J, et al. Adaptive Radiotherapy for Prostate Cancer using Kilovoltage Cone-Beam Computed Tomography: First Clinical Results. Int J Radiat Oncol 2008;70:75–82. doi:10.1016/j.ijrobp.2007.05.046.[12]Sonke J-J, Belderbos J. Adaptive Radiotherapy for Lung Cancer. Semin Radiat Oncol 2010;20:94–106. doi:10.1016/j.semradonc.2009.11.003.[13]Zelefsky MJ, Kollmeier M, Cox B, Fidaleo A, Sperling D, Pei X, et al. Improved clinical outcomes with high-dose image guided radiotherapy compared with non-IGRT for the treatment of clinically localized prostate cancer. Int J Radiat Oncol Biol Phys 2012;84:125–9. doi:10.1016/j.ijrobp.2011.11.047.[14]Bujold A, Craig T, Jaffray D, Dawson LA. Image-Guided Radiotherapy: Has It Influenced Patient Outcomes? Semin Radiat Oncol 2012;22:50–61. doi:10.1016/j.semradonc.2011.09.001.[15]Jaffray DA. Image-guided radiotherapy: from current concept to future perspectives. Nat Rev Clin Oncol 2012;9:688–99. doi:10.1038/nrclinonc.2012.194.[16]Johnson C, Price G, Khalifa J, Faivre-Finn C, Dekker A, Moore C, et al. A method to combine target volume data from 3D and 4D planned thoracic radiotherapy patient cohorts for machine learning applications. Radiother Oncol 2018;126:355–61. doi:10.1016/j.radonc.2017.11.015.[17]Benjamini Y, Hochberg Y. Controlling the False Discovery Rate?: A Practical and Powerful Approach to Multiple Testing. J R Stat Soc Ser B 1995;57:289–300.[18]R Core Team. R: A Language and Environment for Statistical Computing 2016.[19]Ishwaran H, Kogalur UB. Random Forests for Survival, Regression and Classification (RF-SRC) 2016.[20]Ishwaran H, Kogalur U. Random survival forests for R. RNews 2007;7:25–31.[21]Ishwaran H, Kogalur UB, Blackstone EH, Lauer MS. Random survival forests. Ann Appl Stat 2008;2:841–60. doi:10.1214/08-AOAS169.[22]Kassambara A. Survminer: Drawing Survival Curves using “ggplot2” 2016.[23]Therneau TM, Grambsch PM. Modeling Survival Data: Extending the Cox Model. New York: Springer; 2000.[24]Therneau TM, Lumley T. A Package for Survival Analysis in S 2015.[25]Bradley JD, Paulus R, Komaki R, Masters G, Blumenschein G, Schild S, et al. Standard-dose versus high-dose conformal radiotherapy with concurrent and consolidation carboplatin plus paclitaxel with or without cetuximab for patients with stage IIIA or IIIB non-small-cell lung cancer (RTOG 0617): a randomised, two-by-two factorial p. Lancet Oncol 2015:187–99. doi:10.1016/S1470-2045(14)71207-0.[26]Faivre-Finn C. Dose escalation in lung cancer: have we gone full circle? Lancet Oncol 2015;16:125–7. doi:10.1016/S1470-2045(15)70001-X.[27]Mcwilliam A, Kennedy J, Hodgson C, Vasquez E, Faivre-finn C, van Herk M. Radiation dose to heart base linked with poorer survival in lung cancer patients. Eur J Cancer 2017;85:106–13. doi:10.1016/j.ejca.2017.07.053.[28]Persson GF, Scherman Rydh?g J, Josipovic M, Maraldo M V., Nyg?rd L, Costa J, et al. Deep inspiration breath-hold volumetric modulated arc radiotherapy decreases dose to mediastinal structures in locally advanced lung cancer. Acta Oncol (Madr) 2016;55:1053–6. doi:10.3109/0284186X.2016.1142115.Figures and TablesFigure 1Figure 2Figure 3Figure 4Table 1: Patient cohort detailsVariableNSCLC cohort(n = 780)Oesophagus cohort (n = 177)Mean Age (range)70 (31 – 94)69 (41 – 88)Gender Male421 (53.97%)113 (63.84%) Female359 (46.03%)64 (36.16%)ECOG-PS 0111 (14.23%)31 (17.51%) 1364 (46.67%)96 (54.24%) 2198 (25.38%)36 (20.34%) 347 (6.03%)7 (3.95%) 42 (0.26%)0 Missing58 (7.43%)7 (3.95%)Comorbidities 079 (10.13%)35 (19.77%) 1105 (13.46%)29 (16.38%) 289 (11.41%)24 (13.56%) 372 (9.23%)9 (5.09%) Missing435 (55.77%)80 (45.20%)Stage I18 (2.31%)20 (11.30%) II113 (14.49%)41 (23.16%) III520 (66.67%)85 (48.02%) IV38 (4.87)8 (4.52%) Missing91 (11.66%)23 (13.00%)Mean GTV cm37239 Missing104 (13.33%)10 (5.65%)Fractionation 60-66Gy in 30-33#159 (20.38%)- 55Gy in 20#621 (79.62%)63 (35.59%) 50Gy in 25#-104 (58.76%) 41.1Gy in 23#-10 (5.65%)Table 2: Multivariate Cox regression resultsCohortNVariablep-valueHR (low shift group)All780SD of lateral shift (>1.5mm vs ≤1.5mm)<0.0011.405ECOG-PS0.0321.121Age0.0831.008Fractionation<0.0010.954ln(GTV)<0.0011.400Left tumour cases261Mean lateral shift (> 0.1mm vs ≤ 0.1mm)0.0250.723ECOG-PS0.0321.224Age0.4301.007Fractionation0.0440.966ln(GTV)0.0021.263Right tumour cases367Mean lateral shift (> 0.4mm vs ≤ 0.4mm)0.0071.401ECOG-PS0.0941.132Age0.3401.006Fractionation<0.0010.943ln(GTV)<0.0011.457Superior tumour cases566Mean longitudinal shift (> -1.8mm vs ≤ -1.8mm)0.0110.664ECOG-PS0.0201.151Age0.3851.004Fractionation<0.0010.949ln(GTV)<0.0011.371Posterior tumour cases599Mean vertical shift (> -1.2mm vs ≤ -1.2mm)0.0031.379ECOG-PS0.0071.171Age0.2201.006Fractionation<0.0010.953ln(GTV)<0.0011.386All780Vector shift to heart (> 0.0mm vs ≤ 0.0mm)<0.0010.757ECOG-PS0.0091.148Age0.2141.006Fractionation<0.0010.955ln(GTV)<0.0011.405Figure LegendsTable 1: Patient details for each variable in the NSCLC and oesophagus cohorts, giving the number of cases (and corresponding percentage of the whole cohort) in each variable level. Where there is missing data in the variable, this is given as an additional level. Table 2: Details of the split points used to assign the different residual shift metrics to risk categories together with multivariate Cox regression hazard ratios (HR) and p-values for each analysis. HRs associated with shift categories give the hazard of death for patients in the low residual shift group as compared to the high residual shift group i.e. HR of having low SD and low mean shifts.ln(GTV) = natural logarithm of the gross tumour volume.Figure 1: a) Schematic of how the residual IGRT shifts were calculated, by first retrospectively applying the shifts that were over the 5mm action threshold (highlighted) and then using a nearest-neighbour weighting approach to backfill data in fractions where no imaging was used. b) Schematic of how the vector shift to the heart was calculated. Using the centre of mass (CoM) as a representative point for structure position, the vector length between the PTV CoM – taken to be the region of high dose represented here by cross-hairs – and the heart CoM was determined. The residual shifts were then applied to the heart position, and this vector length was re-calculated. The difference between these two values, represented by the two red arrows, determines whether a shift moves the heart towards or away from the high dose region.Figure 2: Plots of the correlation with the vector shift to the heart for a) patient age, b) performance status, c) GTV volume and d) comorbidities (subset of patients with comorbidity scores available, n=345).PCC = Pearson’s Correlation Coefficient; AOV = Analysis of VarianceFigure 3: Multivariate Cox regression survival curves for the (a) left (n = 261) and (b) right (n = 367) cohorts stratified into risk groups using the mean lateral shift split points detailed inTable 2, and (c) the whole cohort (n = 780) stratified on the vector shift towards or away from the heart (split point 0.0mm). Patients with high shifts in the left cohort (those > 0.1mm, meaning that the majority ofshifts will be to the left, equating to moving the heart closer to the high dose region) have worse overall survival (p = 0.025), while patients with high shifts in the right cohort (those > 0.4mm, meaning that the majority of shifts will be to the left, moving the heart away from the high dose region) have improved overall survival (p = 0.007). For the vector shift to the heart, a significant survival difference was observed (p < 0.001).Figure 4: Multivariate Cox regression survival curves for the high and low shift groups, for the oesophagus cancer cohort (n = 177), stratified on the three dimensional residual shift towards or away from the heart (split point 0.0mm). Positive residual shifts represent the high dose region being shifted towards the heart, whilst negative shifts represent the high dose region being shifted away from the heart. A significant survival difference was seen between shifts towards and away from the heart (p = 0.03). ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download