- Email: [email protected]
Results of Patterns of Care Study of Patients Treated with External Beam Radiotherapy for Prostate Cancer in 2004
Proceedings of the 53rd Annual ASTRO Meeting at 12 months (mean = 1.53). Vaginal tightness during dilation was higher than baseline (mean = 1.85) at 12 months post-RT (mean = 2.09). Only 5 pts had mild bleeding with use of vaginal dilators at 12 months. Conclusions: Compliance with the use of a vaginal dilator is reasonable and based on this preliminary analysis may reduce vaginal stenosis; however, the increase in vaginal pain during dilation at 12 months emphasizes the need for ongoing reinforcement of dilator compliance. Further longitudinal studies are needed to understand reasons for non-compliance and develop strategies to increase use of vaginal dilators. Author Disclosure: E. Law: None. N. Kline: None. B. Thom: None. K. Alektiar: None. E. Riedel: None. K. Kaufman: None. G. Cohen: None. J. Kelvin: None. D. Paolilli: None. K. Goodman: None.
2731
A Nomogram for Predicting the Benefit of Adjuvant Therapy for Resected Pancreatic Ductal Adenocarcinoma
C. Bicquart, R. El Youssef, A. R. Wissel, B. C. Sheppard, K. G. Billingsley, C. R. Thomas, S. J. Wang Oregon Health Sciences University, Portland, OR Purpose/Objective(s): Given the poor long-term outcomes of resected pancreatic ductal adenocarcinoma (PDAC), the addition of adjuvant therapy is one avenue to improve survival and local control outcomes. Yet, there is no clear consensus regarding the optimal adjuvant therapy for resected PDAC. Adjuvant treatment can consist of either chemotherapy and/or radiotherapy, and is often decided based on the pathological findings after resection. Identifying which patients will benefit from adjuvant therapy and estimating the magnitude of this benefit remains challenging. The specific aim of this study was to create a decision aid to help make individualized estimates of the potential survival benefit of adjuvant chemotherapy or chemoradiotherapy (CRT) for patients with resected pancreatic cancer. Materials/Methods: Patients who underwent pancreaticoduodenectomy at OHSU between 1996- 2010 were included. Potential covariates were analyzed including age, sex, stage, nodal status, lymphadenectomy, histology, margin status, lymphovascular space invasion, and perineural invasion. The primary endpoint was overall survival after surgery alone, with adjuvant chemo alone, or with adjuvant CRT. Several types of multivariate survival regression models were constructed and compared—Cox proportional hazards, Weibull, exponential, log-logistic, and lognormal. Forward stepwise variable selection was used to select the final covariates included in the model. Model performance was compared using the Akaike Information Criterion (AIC). The best performing model was internally validated for both discrimination (concordance index) and calibration (calibration curve) using bootstrap resampling. Results: 179 patients met the inclusion criteria for the study. At the time of analysis, 82% of patients had already expired. The median overall survival for the entire cohort was 16.4 months. The median overall survival for pts who received no adjuvant therapy, adjuvant chemo alone, and adjuvant CRT was 15.2, 12.6, and 20.8 months, respectively. When models were compared using the AIC, the log-logistic and lognormal survival models showed the best performance. A web browser-based nomogram was built from the lognormal model that can make individualized estimates of survival benefit from adjuvant therapy. The concordance index of this model was 0.66. The model predicts that nodal status, margin status, and perineural invasion are the most important factors predicting outcome and patients who had positive nodes, positive margins, or perineural invasion yielded the largest benefit from adjuvant CRT. Conclusions: An online nomogram built from a parametric survival model can be used as a decision aid to predict and quantify the expected benefit from adjuvant chemo or CRT for pancreatic cancer. Author Disclosure: C. Bicquart: None. R. El Youssef: None. A.R. Wissel: None. B.C. Sheppard: None. K.G. Billingsley: None. C.R. Thomas: None. S.J. Wang: None.
2732
Results of Patterns of Care Study of Patients Treated with External Beam Radiotherapy for Prostate Cancer in 2004
D. Wang1, A. Ho2, X. Wu3, A. Hamilton4, M. Goodman5, S. Fleming6, C. Rao7, R. German8, J. Owen2 1 Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, WI, 2Clinical Research Center, American College of Radiology, Philadelphia, PA, 3Louisiana Tumor Registry, LSU Health Sci Ctr, New Orleans, LA, 4Keck School of Medicine, University of Southern California, Los Angeles, CA, 5Dept Epidemiol, Rollins Sch Pub Health. Emory University, Atlanta, GA, 6University of Kentucky College of Pub Health, Lexington, KY, 7NC Central Cancer Registry, Raleigh, NC, 8 Division of Cancer Prevention and Control, CDC, Atlanta, GA Purpose/Objective(s): To report results of a 2004 Prostate Cancer Data Quality and Patterns of Care Study (POC BP) that expanded data routinely collected by the state cancer registry for prostate cancer patients treated with radiotherapy (RT) in seven states (CA, GA, KY, LA, MN, NC, WI) and to compare the results with the 1999 Patterns of Care Study in Radiation Oncology (PCS). Materials/Methods: 9017 prostate cancer patient records were abstracted from stratified sample of newly diagnosed cases in 2004 from the above states. Clinical practice trends were determined with weighted analyses according to recurrence risk groups of NCCN 2002 Guidelines and compared with results of the previous national PCS including use of androgen deprivation therapy (ADT) in combination with EBRT. Parameters used to characterize the recurrence risk groups include pretreatment PSA, T stage and Gleason score. Patients were categorized as low risk (LR) if he had T1-T2a, GS 2 - 6, and PSA\10 ng/ml. Intermediate risk (IR) were those with T2b-T2c or GS = 7 or PSA 10 - 20 ng/ml, and high/very high risk (HR) were those with T3a-T4 or GS 8 - 10 or PSA.20 ng/ml. Results: Of the 3548 men (weighted sample size = 10112) with non-metastatic disease who had RT, 41.0% were treated with brachytherapy with or without external beam radiotherapy (EBRT), 57.8% were treated with EBRT only, and 1.2% unknown modality. Compared with the 1999 PCS, the number of patients in the POC BP survey treated with brachytherapy was higher (41.0% vs. 36%). The EBRT only group, which included 28.6% LR, 42.5% IR, and 28.9% HR groups, is the subject of the remaining
S561
I. J. Radiation Oncology d Biology d Physics
S562
Volume 81, Number 2, Supplement, 2011
findings. Intensity modulated radiation treatment (IMRT) was used to treat 31.2% of prostate cancer treated with EBRT only. Percent of ADT use in LR, IR and HR groups was 30%, 51.2%, 83.7%, respectively (p\0.0001). A multivariate logistic regression analysis revealed a statistically significant increased likelihood (odds ratio (OR) = 2.6 and 12.7, p\0.0001) of IR-HR groups respectively being treated with ADT in conjunction with EBRT compared with LR group. Percent of ADT combined with EBRT for localized prostate cancer is similar to the 1999 PCS (54.5% and 51% respectively). Conclusions: Compared with the 1999 PCS, use of brachytherapy was higher. Use of ADT combined with EBRT remained high in treatment of prostate cancer with all risk groups. Frequent use of IMRT was discovered in this 2004 prostate cancer survey. Acknowledgment: Supported by the Centers for Disease Control and Prevention 1-U01-DP000260, 1-U01-DP000258, 1-U01DP000253, 1-U01-DP000259, 1-U01-DP000261, 1-U01-DP000251, 1-U01-DP000264. Author Disclosure: D. Wang: None. A. Ho: None. X. Wu: None. A. Hamilton: None. M. Goodman: None. S. Fleming: None. C. Rao: None. R. German: None. J. Owen: None.
2733
Retrospective Exploration of the Impact of Quality Assurance Rounds in a Canadian Radiotherapy Department
S. Lefresne R. Olson BC Cancer Agency, Vancouver, BC, Canada Purpose/Objective(s): Quality assurance (QA) programs are necessary in radiation oncology to identify potential errors or inconsistencies that may compromise the quality of patient care. The prescription of radiation doses and the delineation of tumor volumes are two well-documented and important sources of inconsistency. Some cancer centers have implemented QA rounds in order to identify and manage these inconsistencies. The purpose of this study was to evaluate the outcome of the weekly QA rounds at a single institution. Materials/Methods: QA round records, from the dates of 2004 - 2010, were obtained from an urban cancer center. During these rounds, radiotherapy plans were randomly selected, evaluated by a multidisciplinary team, and assigned a recommendation of either ‘no changes’ (NC), ‘minor changes’ (MC), or ‘significant changes’ (SC) needed. The relationship between recommendations for each plan, tumor site, and mean years of experience of the radiation oncologist (RO) were explored. Results: A total of 1247 plans and 20 ROs were evaluated during the study period. Eight-five percent of the plans were curative in intent. The most common tumor sites were breast (31.2%), genitourinary (18.8%) and gastrointestinal (13.1%). Overall, 7.0% of plans received a recommendation for either MC or SC. Among the radical plans, the mean years of experience of RO with a recommendation of NC, MC, and SC were 19.6, 19.8, and 13.7, respectively (p = 0.065). On exploratory analysis, the mean RO years of experience were significantly lower for plans with SC versus NC recommendations (p = 0.020). The three most common sites to receive a recommendation for change (MC or SC) were gastrointestinal (13.8%), lung (12.5%) and lymphoma (7.9%). Conclusions: In this study, recommendations for minor or significant changes were relatively common (7.0%), with plans for gastrointestinal, lung and lymphoma tumor sites most often impacted. In addition, there was an inverse relationship between years of experience and a recommendation for change in the radiotherapy plan. This finding suggests that QA rounds may be particularly beneficial for radiation oncologists who are new to practice. Author Disclosure: S. Lefresne: None. R. Olson: None.
2734
Failure Modes and Effects Analysis of an Electronic Quality-Checklist Process Map in Radiation Medicine: Has it Made a Difference?
A. Kapur, L. Potters, R. Sharma, L. Lee, Y. Cao, P. Zuvic, N. Adair, E. Montchal, L. Vijeh, L. B. Mallalieu North Shore Long Island Jewish Health System, Department of Radiation Medicine, New Hyde Park, NY Purpose/Objective(s): The purpose of this work was to stratify procedures in a multidisciplinary electronic quality checklist (QCL) into a patient-safety risk pareto chart using process failure modes and effects analysis (FMEA) and to quantify likelihood-of-occurrence risk reductions using measured delays with the advent of a ‘‘No-Fly’’ process interlock safety control policy. Materials/Methods: An electronic QCL process map (QPM) in use at our facility for several months was cast in the framework of an FMEA spreadsheet. Risk scoring guides were established for the three components of risk - severity, likelihood-of-occurrence and detectability using logarithmic risk scales ranging from 1 to 10 in ascending risk order. A risk priority number (RPN) was computed as the product of the three risk elements and the QPM sorted in descending order to arrange QCL procedures by high, intermediate and low risk levels. The exercise was conducted by oncologists, physicists, dosimetrists, nurses, therapists and informatics staff, with 2 - 4 members in each group bracketing the hierarchical organizational spread within each group. Procedural failure modes were identified by extracting actual events registered in our internally developed aspects-of-care event reporting database, group discussions and national event registries. Causes of failure modes were established by prior root-causeanalysis. Effects of failure modes were streamlined to the context of patient safety. Independent responses were amalgamated and average scores computed. A ‘‘No-Fly’’ policy (NFP) was introduced to provide hard stops for highest risk procedures identified with the FMEA. An independent audit of likelihood-of-occurrence risks was subsequently performed using delay statistics extracted directly from our Oncology-Information-System before and after the introduction of the safety policy. Results: The highest RPN procedures included contours, prescription, treatment plan completion, MD plan approvals, IMRT QA, 2nd physics checks, pathology review, patient consent, laterality and 1st day physics checks. On average, for the high risk procedures the fraction of delays and mean slip days improved by factors of 1.6 and 2.1 respectively with the NFP. Conclusions: The FMEA while semi-quantitative led to process control policies that showed measurable differences in risk reduction using slip analysis. Delays were chosen as surrogates for likelihood-of-occurrence risks as they are measurable, suggestive of procedural failures and progenitors for hastened downstream tasks that potentially culminate in risk-expedited care. Author Disclosure: A. Kapur: None. L. Potters: None. R. Sharma: None. L. Lee: None. Y. Cao: None. P. Zuvic: None. N. Adair: None. E. Montchal: None. L. Vijeh: None. L.B. Mallalieu: None.
2731
A Nomogram for Predicting the Benefit of Adjuvant Therapy for Resected Pancreatic Ductal Adenocarcinoma
C. Bicquart, R. El Youssef, A. R. Wissel, B. C. Sheppard, K. G. Billingsley, C. R. Thomas, S. J. Wang Oregon Health Sciences University, Portland, OR Purpose/Objective(s): Given the poor long-term outcomes of resected pancreatic ductal adenocarcinoma (PDAC), the addition of adjuvant therapy is one avenue to improve survival and local control outcomes. Yet, there is no clear consensus regarding the optimal adjuvant therapy for resected PDAC. Adjuvant treatment can consist of either chemotherapy and/or radiotherapy, and is often decided based on the pathological findings after resection. Identifying which patients will benefit from adjuvant therapy and estimating the magnitude of this benefit remains challenging. The specific aim of this study was to create a decision aid to help make individualized estimates of the potential survival benefit of adjuvant chemotherapy or chemoradiotherapy (CRT) for patients with resected pancreatic cancer. Materials/Methods: Patients who underwent pancreaticoduodenectomy at OHSU between 1996- 2010 were included. Potential covariates were analyzed including age, sex, stage, nodal status, lymphadenectomy, histology, margin status, lymphovascular space invasion, and perineural invasion. The primary endpoint was overall survival after surgery alone, with adjuvant chemo alone, or with adjuvant CRT. Several types of multivariate survival regression models were constructed and compared—Cox proportional hazards, Weibull, exponential, log-logistic, and lognormal. Forward stepwise variable selection was used to select the final covariates included in the model. Model performance was compared using the Akaike Information Criterion (AIC). The best performing model was internally validated for both discrimination (concordance index) and calibration (calibration curve) using bootstrap resampling. Results: 179 patients met the inclusion criteria for the study. At the time of analysis, 82% of patients had already expired. The median overall survival for the entire cohort was 16.4 months. The median overall survival for pts who received no adjuvant therapy, adjuvant chemo alone, and adjuvant CRT was 15.2, 12.6, and 20.8 months, respectively. When models were compared using the AIC, the log-logistic and lognormal survival models showed the best performance. A web browser-based nomogram was built from the lognormal model that can make individualized estimates of survival benefit from adjuvant therapy. The concordance index of this model was 0.66. The model predicts that nodal status, margin status, and perineural invasion are the most important factors predicting outcome and patients who had positive nodes, positive margins, or perineural invasion yielded the largest benefit from adjuvant CRT. Conclusions: An online nomogram built from a parametric survival model can be used as a decision aid to predict and quantify the expected benefit from adjuvant chemo or CRT for pancreatic cancer. Author Disclosure: C. Bicquart: None. R. El Youssef: None. A.R. Wissel: None. B.C. Sheppard: None. K.G. Billingsley: None. C.R. Thomas: None. S.J. Wang: None.
2732
Results of Patterns of Care Study of Patients Treated with External Beam Radiotherapy for Prostate Cancer in 2004
D. Wang1, A. Ho2, X. Wu3, A. Hamilton4, M. Goodman5, S. Fleming6, C. Rao7, R. German8, J. Owen2 1 Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, WI, 2Clinical Research Center, American College of Radiology, Philadelphia, PA, 3Louisiana Tumor Registry, LSU Health Sci Ctr, New Orleans, LA, 4Keck School of Medicine, University of Southern California, Los Angeles, CA, 5Dept Epidemiol, Rollins Sch Pub Health. Emory University, Atlanta, GA, 6University of Kentucky College of Pub Health, Lexington, KY, 7NC Central Cancer Registry, Raleigh, NC, 8 Division of Cancer Prevention and Control, CDC, Atlanta, GA Purpose/Objective(s): To report results of a 2004 Prostate Cancer Data Quality and Patterns of Care Study (POC BP) that expanded data routinely collected by the state cancer registry for prostate cancer patients treated with radiotherapy (RT) in seven states (CA, GA, KY, LA, MN, NC, WI) and to compare the results with the 1999 Patterns of Care Study in Radiation Oncology (PCS). Materials/Methods: 9017 prostate cancer patient records were abstracted from stratified sample of newly diagnosed cases in 2004 from the above states. Clinical practice trends were determined with weighted analyses according to recurrence risk groups of NCCN 2002 Guidelines and compared with results of the previous national PCS including use of androgen deprivation therapy (ADT) in combination with EBRT. Parameters used to characterize the recurrence risk groups include pretreatment PSA, T stage and Gleason score. Patients were categorized as low risk (LR) if he had T1-T2a, GS 2 - 6, and PSA\10 ng/ml. Intermediate risk (IR) were those with T2b-T2c or GS = 7 or PSA 10 - 20 ng/ml, and high/very high risk (HR) were those with T3a-T4 or GS 8 - 10 or PSA.20 ng/ml. Results: Of the 3548 men (weighted sample size = 10112) with non-metastatic disease who had RT, 41.0% were treated with brachytherapy with or without external beam radiotherapy (EBRT), 57.8% were treated with EBRT only, and 1.2% unknown modality. Compared with the 1999 PCS, the number of patients in the POC BP survey treated with brachytherapy was higher (41.0% vs. 36%). The EBRT only group, which included 28.6% LR, 42.5% IR, and 28.9% HR groups, is the subject of the remaining
S561
I. J. Radiation Oncology d Biology d Physics
S562
Volume 81, Number 2, Supplement, 2011
findings. Intensity modulated radiation treatment (IMRT) was used to treat 31.2% of prostate cancer treated with EBRT only. Percent of ADT use in LR, IR and HR groups was 30%, 51.2%, 83.7%, respectively (p\0.0001). A multivariate logistic regression analysis revealed a statistically significant increased likelihood (odds ratio (OR) = 2.6 and 12.7, p\0.0001) of IR-HR groups respectively being treated with ADT in conjunction with EBRT compared with LR group. Percent of ADT combined with EBRT for localized prostate cancer is similar to the 1999 PCS (54.5% and 51% respectively). Conclusions: Compared with the 1999 PCS, use of brachytherapy was higher. Use of ADT combined with EBRT remained high in treatment of prostate cancer with all risk groups. Frequent use of IMRT was discovered in this 2004 prostate cancer survey. Acknowledgment: Supported by the Centers for Disease Control and Prevention 1-U01-DP000260, 1-U01-DP000258, 1-U01DP000253, 1-U01-DP000259, 1-U01-DP000261, 1-U01-DP000251, 1-U01-DP000264. Author Disclosure: D. Wang: None. A. Ho: None. X. Wu: None. A. Hamilton: None. M. Goodman: None. S. Fleming: None. C. Rao: None. R. German: None. J. Owen: None.
2733
Retrospective Exploration of the Impact of Quality Assurance Rounds in a Canadian Radiotherapy Department
S. Lefresne R. Olson BC Cancer Agency, Vancouver, BC, Canada Purpose/Objective(s): Quality assurance (QA) programs are necessary in radiation oncology to identify potential errors or inconsistencies that may compromise the quality of patient care. The prescription of radiation doses and the delineation of tumor volumes are two well-documented and important sources of inconsistency. Some cancer centers have implemented QA rounds in order to identify and manage these inconsistencies. The purpose of this study was to evaluate the outcome of the weekly QA rounds at a single institution. Materials/Methods: QA round records, from the dates of 2004 - 2010, were obtained from an urban cancer center. During these rounds, radiotherapy plans were randomly selected, evaluated by a multidisciplinary team, and assigned a recommendation of either ‘no changes’ (NC), ‘minor changes’ (MC), or ‘significant changes’ (SC) needed. The relationship between recommendations for each plan, tumor site, and mean years of experience of the radiation oncologist (RO) were explored. Results: A total of 1247 plans and 20 ROs were evaluated during the study period. Eight-five percent of the plans were curative in intent. The most common tumor sites were breast (31.2%), genitourinary (18.8%) and gastrointestinal (13.1%). Overall, 7.0% of plans received a recommendation for either MC or SC. Among the radical plans, the mean years of experience of RO with a recommendation of NC, MC, and SC were 19.6, 19.8, and 13.7, respectively (p = 0.065). On exploratory analysis, the mean RO years of experience were significantly lower for plans with SC versus NC recommendations (p = 0.020). The three most common sites to receive a recommendation for change (MC or SC) were gastrointestinal (13.8%), lung (12.5%) and lymphoma (7.9%). Conclusions: In this study, recommendations for minor or significant changes were relatively common (7.0%), with plans for gastrointestinal, lung and lymphoma tumor sites most often impacted. In addition, there was an inverse relationship between years of experience and a recommendation for change in the radiotherapy plan. This finding suggests that QA rounds may be particularly beneficial for radiation oncologists who are new to practice. Author Disclosure: S. Lefresne: None. R. Olson: None.
2734
Failure Modes and Effects Analysis of an Electronic Quality-Checklist Process Map in Radiation Medicine: Has it Made a Difference?
A. Kapur, L. Potters, R. Sharma, L. Lee, Y. Cao, P. Zuvic, N. Adair, E. Montchal, L. Vijeh, L. B. Mallalieu North Shore Long Island Jewish Health System, Department of Radiation Medicine, New Hyde Park, NY Purpose/Objective(s): The purpose of this work was to stratify procedures in a multidisciplinary electronic quality checklist (QCL) into a patient-safety risk pareto chart using process failure modes and effects analysis (FMEA) and to quantify likelihood-of-occurrence risk reductions using measured delays with the advent of a ‘‘No-Fly’’ process interlock safety control policy. Materials/Methods: An electronic QCL process map (QPM) in use at our facility for several months was cast in the framework of an FMEA spreadsheet. Risk scoring guides were established for the three components of risk - severity, likelihood-of-occurrence and detectability using logarithmic risk scales ranging from 1 to 10 in ascending risk order. A risk priority number (RPN) was computed as the product of the three risk elements and the QPM sorted in descending order to arrange QCL procedures by high, intermediate and low risk levels. The exercise was conducted by oncologists, physicists, dosimetrists, nurses, therapists and informatics staff, with 2 - 4 members in each group bracketing the hierarchical organizational spread within each group. Procedural failure modes were identified by extracting actual events registered in our internally developed aspects-of-care event reporting database, group discussions and national event registries. Causes of failure modes were established by prior root-causeanalysis. Effects of failure modes were streamlined to the context of patient safety. Independent responses were amalgamated and average scores computed. A ‘‘No-Fly’’ policy (NFP) was introduced to provide hard stops for highest risk procedures identified with the FMEA. An independent audit of likelihood-of-occurrence risks was subsequently performed using delay statistics extracted directly from our Oncology-Information-System before and after the introduction of the safety policy. Results: The highest RPN procedures included contours, prescription, treatment plan completion, MD plan approvals, IMRT QA, 2nd physics checks, pathology review, patient consent, laterality and 1st day physics checks. On average, for the high risk procedures the fraction of delays and mean slip days improved by factors of 1.6 and 2.1 respectively with the NFP. Conclusions: The FMEA while semi-quantitative led to process control policies that showed measurable differences in risk reduction using slip analysis. Delays were chosen as surrogates for likelihood-of-occurrence risks as they are measurable, suggestive of procedural failures and progenitors for hastened downstream tasks that potentially culminate in risk-expedited care. Author Disclosure: A. Kapur: None. L. Potters: None. R. Sharma: None. L. Lee: None. Y. Cao: None. P. Zuvic: None. N. Adair: None. E. Montchal: None. L. Vijeh: None. L.B. Mallalieu: None.