Development of a Semi-Automated Chart Review for Assessing the Development of Radiation Pneumonitis: using Natural Language Processing (Preprint)

BACKGROUND Health research frequently requires manual chart review to identify patients for the study-specific cohort and examine their clinical outcomes. Manual chart review is a labour-intensive process requiring significant time investment for clinical researchers. OBJECTIVE This study aimed to evaluate the feasibility and accuracy of an assisted chart review program, using an in-house natural language processing (NLP) program, to identify patients who developed radiation pneumonitis (RP) after receiving curative radiotherapy. METHODS A retrospective manual chart review was completed for patients who received curative radiotherapy for stage II-III lung cancer from January 1, 2013 to December 31, 2015 at BC Cancer Kelowna. In the manual chart review, RP diagnosis and grading were recorded using Common Terminology Criteria for Adverse Events (CTCAE) v5.0. From the charts of 50 sample patients, a total of 1413 clinical documents were extracted for review from the Cancer Agency Information System (CAIS). The NLP program was built using the Natural Language Toolkit Python platform. Python version 3.7.2. was used to run the NLP program. The output of the NLP program is a list of the full sentences containing the key terms, the document ID’s and dates from which these sentences were extracted. The result from the manual review was used as the gold standard in this study, with which the result of the NLP program was compared. RESULTS Twenty-five out of the 50 sample patients developed RP grade 1 or greater; the NLP program was able to ascertain 23 out of these 25 patients (sensitivity = 0.92, 95%CI:0.74-0.99; specificity = 0.36, 95%CI:0.18-0.57). Furthermore, the NLP program was able to correctly identify all 9 patients with RP grade 2 or greater, which are patients with clinically significant symptoms (sensitivity = 1.0, 95%CI: 0.66-1.0; specificity = 0.27, 95%CI:0.14-0.43). The NLP program was useful in distinguishing patients with RP from those without RP. The NLP program in this study would avoid unnecessary manual review of 22% of the sample patients (n=11), as these patients were identified as RP grade 0 and will not require further manual review in subsequent studies. CONCLUSIONS This feasibility study showed that the NLP program was able to assist with the identification of patients who developed RP after curative radiotherapy. The NLP program streamlines the manual chart review further by identifying key sentences of interest. This work has a potential to improve future clinical research, as the NLP program shows promise in performing chart review in a more time efficient manner, compared to the traditional labor-intensive manual chart review.

The role of leadership in driving change in cancer survival outcomes in seven high income countries

Background There is well-established variation in cancer survival across high-income countries with seemingly-similar health systems. There is much research on reasons for these differences, but the role of leadership has been under-researched despite being one of the WHO 'building blocks' that underpin a functioning health system. Leadership is variously defined as including governance, stewardship, responsibility and accountability. As part of the International Cancer Benchmarking Partnership, this study looked at these diverse aspects of leadership to identify drivers of change and improvement across a range of high-income countries. Methods Cancer strategy documents were analysed from 22 jurisdictions: Australia (3 states), Canada (10 provinces), Denmark, Ireland, New Zealand, Norway, and UK (4 countries). Key informants in 15 of these jurisdictions, representing a range of stakeholders, were recruited through a combination of purposive and 'snowball' strategies, and invited to participate in semi-structured interviews. Documents and interview transcripts were analysed using a thematic approach. Results Different facets of leadership emerged: diffused vs centralised (including the central role of a cancer agency in some places); national, regional and local leadership structures; links between primary and secondary care. This study, however, demonstrated a central role for sustained leadership and political commitment, crucial for initiating and maintaining progress, as was a coherent vision that supported the implementation of national policies locally. Clinical leadership emerged as vital for translating policy into action. Conclusions Improving cancer outcomes is challenging and complex but is unlikely to be achieved without effective leadership and sustained political commitment that can create effective co-ordination of care. Key messages Different facets of leadership of the cancer care system emerged as important when exploring the reasons for variations in cancer outcomes in high-income countries. The persistence of these variations is unacceptable. Change will require political commitment and effective leadership, especially by clinicians.

Sequenztherapie beim nicht oder mild symptomatischen metastasierten kastrationsresistenten Prostatakarzinom

ZusammenfassungBeim nicht oder mild symptomatischen metastasierten kastrationsresistenten Prostatakarzinom (mCRPC) können sowohl die modernen antihormonellen Wirkstoffe Abirateronacetat plus Prednison/Prednisolon (Abirateron/P) und Enzalutamid als auch eine Chemotherapie mit Docetaxel zum Einsatz kommen. Da die Behandlung mit Docetaxel teils ausgeprägte Nebenwirkungen haben kann, ist eine wichtige Therapiealternative die insgesamt besser verträgliche antihormonelle Therapie. Neben der Wirksamkeit und Verträglichkeit kann bei der Wahl der mCRPC-Erstlinientherapie zudem der Einfluss auf die Wirksamkeit von Folgetherapien eine wichtige Rolle spielen. Bisher lagen nur retrospektive Studien zum sequenziellen Einsatz von Abirateron/P und Enzalutamid bei dieser Indikation vor. In einer prospektiven, randomisierten Phase-II-Studie der British Columbia Cancer Agency (BCCA) wurden Abirateron/P und Enzalutamid nun erstmals bez. ihrer Wirksamkeit und Verträglichkeit in der Erstlinientherapie des nicht oder mild symptomatischen mCRPC sowie der Folgetherapie miteinander verglichen.

A method to audit and score implementation of knowledge translation (KT) interventions in large health regions – an observational pilot study using rectal cancer surgery in Ontario

Background: Across Ontario, since the year 2006 various knowledge translation (KT) interventions designed to improve the quality of rectal cancer surgery have been implemented by the provincial cancer agency or by individual researchers. Ontario is divided administratively into 14 health regions. We piloted a method to audit and score for each region of the province the KT interventions implemented to improve the quality of rectal cancer surgery. Methods: We interviewed stakeholders to audit KT interventions used in respective regions over years 2006 to 2014. Results were summarized into narrative and visual forms. Using a modified Delphi approach, KT experts reviewed these data and then, for each region, scored implementation of KT interventions using a 20-item KT Signature Assessment Tool. Scores could range from 20-100 with higher scores commensurate with greater KT intervention implementation. Results: There were thirty interviews. KT experts produced scores for each region that were bimodally distributed, with an average score for 2 regions of 78 (range 73-83) and for 12 regions of 30.5 (range 22-38). Conclusion: Our methods efficiently identified two groups with similar KT Signature scores. Two regions had relatively high scores reflecting numerous KT interventions and the use of sustained iterative approaches in addition to those encouraged by the provincial cancer agency, while 12 regions had relatively low scores reflecting minimal activities outside of those encouraged by the provincial cancer agency. These groupings will be used for future comparative quantitative analyses to help determine if higher KT signature scores correlate with improved measures for quality of rectal cancer surgery.

Impact of surveillance among patients with resected pancreatic cancer following adjuvant chemotherapy.

e16795 Background: Pancreatic adenocarcinoma carries a poor prognosis and high risk of recurrence even after surgery and adjuvant chemotherapy (AC). Guidelines recommend against routine surveillance imaging due to lack of evidence supporting a survival benefit. With current first-line chemotherapy options, it is unclear whether surveillance scans allow for early detection of asymptomatic disease and therefore an opportunity to offer fit patients chemotherapy. We describe the patterns of surveillance in patients followed at a Canadian provincial cancer agency and determine whether routine imaging after AC is associated with receipt of palliative chemotherapy (PC). Methods: A retrospective review was completed to identify patients treated at British Columbia (BC) Cancer centres between January 1, 2010 and December 31, 2016 who had undergone curative intent resection and received at least one cycle of AC. Baseline characteristics, number of scans done after completing AC to recurrence, and PC were collected. Logistic regression analysis was performed. Results: A total of 151 patients followed at BC Cancer were identified. Patients who recurred within 28 days after AC were excluded, leaving 142 patients, of which 115 patients had recurrence (81%). We defined 2 cohorts based on number of scans done between completion of AC and recurrence: those with 0-1 scans were “symptomatic” recurrences (22 patients, median age 68y, 64% female, and 91% node-positive) and those with > 1 scan were “surveillance” recurrences (93 patients, median age 64y, 43% female, and 81% node-positive). Patients who underwent surveillance scans were more likely to receive PC at time of recurrence, though statistical significance was not reached (OR 2.11, 95% CI 0.75-6.58, p = 0.17). Conclusions: Despite guidelines, the majority of patients treated in BC underwent surveillance imaging. Within the limits of our sample size, we demonstrated a trend towards increased likelihood of receiving PC in patients who receive surveillance scans following AC. With efficacious PC options available, studies to determine whether receipt of PC in asymptomatic recurrences detected on imaging translates into improved survival and/or quality of life are warranted.

A method to audit and score implementation of knowledge translation (KT) interventions in large health regions – an observational pilot study using rectal cancer surgery in Ontario

Background: Across Ontario, since the year 2006 various knowledge translation (KT) interventions designed to improve the quality of rectal cancer surgery have been implemented by the provincial cancer agency or by individual researchers. Ontario is divided administratively into 14 health regions. We piloted a method to audit and score for each region of the province the KT interventions implemented to improve the quality of rectal cancer surgery. Methods: We interviewed stakeholders to audit KT interventions used in respective regions over years 2006 to 2014. Results were summarized into narrative and visual forms. KT experts reviewed these data and then, for each region, scored implementation of KT interventions using a 20-item KT Signature Assessment Tool. Scores could range from 20-100 with higher scores commensurate with greater KT intervention implementation. Results: There were thirty interviews. KT experts produced scores for each region that were bimodally distributed, with an average score for 2 regions of 78 (range 73-83) and for 12 regions of 30.5 (range 22-38). Conclusion: Our methods efficiently identified two groups with similar KT Signature scores. Two regions had relatively high scores reflecting numerous KT interventions and the use of sustained iterative approaches in addition to those encouraged by the provincial cancer agency, while 12 regions had relatively low scores reflecting minimal activities. These groupings will be used for future comparative quantitative analyses to help determine if higher KT signature scores correlate with improved measures for region-level quality of rectal cancer surgery.

A method to audit and score implementation of knowledge translation (KT) interventions in large health regions – an observational pilot study using rectal cancer surgery in Ontario

Background: Across Ontario, since the year 2006 various knowledge translation (KT) interventions designed to improve the quality of rectal cancer surgery have been implemented by the provincial cancer agency or by individual researchers. Ontario is divided administratively into 14 health regions. We piloted a method to audit and score for each region of the province KT interventions implemented to improve the quality of rectal cancer surgery. Methods: We interviewed stakeholders to audit KT interventions used in respective regions over years 2006 to 2014. Results were summarized into narrative and visual forms. KT experts reviewed these data and then, for each region, scored KT intervention implementation using a 20-item KT Signature Assessment Tool. Scores could range from 20-100 with higher scores commensurate with greater KT intervention implementation. Results: There were thirty interviews. KT experts produced scores for each region that were bimodally distributed, with an average score for 2 regions of 78 (range 73-83) and for 12 regions of 30.5 (range 22-38). Conclusion: Our methods efficiently identified two groups with similar KT Signature scores. Two regions had relatively high scores reflecting numerous KT interventions and the use of sustained iterative approaches in addition to those encouraged by the provincial cancer agency, while 12 regions had relatively low scores reflecting minimal activities. These groupings will be used for future comparative quantitative analyses to help determine if higher KT signature scores correlate with improved measures for region-level quality of rectal cancer surgery.

A method to audit and score implementation of knowledge translation (KT) interventions in large health regions – an observational pilot study using rectal cancer surgery in Ontario

Background: Across Ontario, since the year 2006 various knowledge translation (KT) interventions designed to improve the quality of rectal cancer surgery have been implemented by the provincial cancer agency or by individual researchers. Ontario is divided administratively into 14 health regions. We piloted a method to audit and score for each region of the province KT interventions implemented to improve the quality of rectal cancer surgery. Methods: We interviewed stakeholders to audit KT interventions used in respective regions over years 2006 to 2014. Results were summarized into narrative and visual forms. KT experts reviewed these data and then, for each region, scored KT intervention implementation using a 20-item KT Signature Assessment Tool. Scores could range from 20-100 with higher scores commensurate with greater KT intervention implementation. Results: There were thirty interviews. KT experts produced scores for each region that were bimodally distributed, with an average score for 2 regions of 78 (range 73-83) and for 12 regions of 30.5 (range 22-38). Conclusion: Our methods efficiently identified two groups with similar KT Signature scores. Two regions had relatively high scores reflecting numerous KT interventions and the use of sustained iterative approaches in addition to those encouraged by the provincial cancer agency, while 12 regions had relatively low scores reflecting minimal activities. These groupings will be used for future comparative quantitative analyses to help determine if higher KT signature scores correlate with improved measures for region-level quality of rectal cancer surgery.

Impact of surveillance among patients with resected pancreatic cancer following adjuvant chemotherapy.

768 Background: Pancreatic adenocarcinoma carries a poor prognosis and high risk of recurrence even after surgery and adjuvant chemotherapy (AC). Guidelines recommend against routine surveillance imaging due to lack of evidence supporting a survival benefit. With current first-line chemotherapy options, it is unclear whether surveillance scans allow for early detection of asymptomatic disease and therefore an opportunity to offer fit patients chemotherapy. We describe the patterns of surveillance in patients followed at a Canadian provincial cancer agency and determine whether routine imaging after AC is associated with receipt of palliative chemotherapy (PC). Methods: A retrospective review was completed to identify patients treated at British Columbia (BC) Cancer centres between January 1, 2010 and December 31, 2016 who had undergone curative intent resection and received at least one cycle of AC. Baseline characteristics, number of scans done after completing AC to recurrence, and PC were collected. Logistic regression analysis was performed. Results: A total of 151 patients followed at BC Cancer were identified. Patients who recurred within 28 days after AC were excluded, leaving 142 patients, of which 115 patients had recurrence. We defined 2 cohorts based on number of scans done between completion of AC and recurrence: those with 0-1 scans were “symptomatic” recurrences (22 patients, median age 68y, 64% female, and 91% node-positive) and those with > 1 scan were “surveillance” recurrences (93 patients, median age 64y, 43% female, and 81% node-positive). Patients who underwent surveillance scans were more likely to receive PC at time of recurrence, though statistical significance was not reached (OR 2.11, 95% CI 0.75-6.58, p = 0.17). Conclusions: Despite guidelines, the majority of patients treated in BC underwent surveillance imaging. Within the limits of our sample size, we demonstrated a trend towards increased likelihood of receiving PC in patients who receive surveillance scans following AC. With efficacious PC options available, studies to determine whether receipt of PC in asymptomatic recurrences detected on imaging translates into improved survival and/or quality of life are warranted.

INNV-34. PATTERNS OF FAILURE AND DEVELOPMENT OF A NOVEL PROGNOSTIC SCORING SYSTEM IN ELDERLY PATIENTS WITH GLIOBLASTOMA – FOLLOW UP ON 10 YEAR ANALYSIS OF THE BC CANCER AGENCY POPULATION

PURPOSE Glioblastomas (GBM) are the most common primary brain tumour recurring in most patients despite maximal management. Patient selection for appropriate treatment modality remains challenging resulting in heterogeneity in management. We examined the patterns of failure and developed a scoring system for patient stratification to optimise clinical decision making. METHODS 822 adults (BC Cancer Agency registry) diagnosed 2005–2015 age ≥60 with histologically confirmed GBM ICD-O-3 codes (9440/3, 9441/3, 9442/3) were reviewed. Univariate and Kaplan-Meier analysis were performed. Performance status (PS), age and resection status were assigned a score, cummulative maximal (favorable) score of 10 and minimum (unfavorable) score of 3. Patterns of failure were further analysed in the subset of patients with radiographic follow-up. RESULTS PS score of 3(KPS >80, ECOG 0/1), 2 (KPS 60–70, ECOG 2), 1 (KPS < 60, ECOG 3/4) (median OS 11, 6, 3 months respectively), age score and resection status were prognostic for OS with PS resulting in the most significant curve separation (p< 0.0001). Biopsy as compared to STR/GTR resulted in poorer OS in patients over 70 (age score 1/2) but had less impact in patients younger than 70 (age scores 3/4). The median OS for cumulative scores of 9/10 (123 patients), 7/8 (286 patients), 5/6 (313 patients), and 3/4 (55 patients) were 14, 8, 4 and 2 months respectively (p< 0.0001) allowing for stratification into 4 prognostic groups. 133 patients had >3 MRIs following diagnosis allowing for clinical and radiographic analysis of progression. Clinical/radiographic progression occurred within 3 months (29%/45%), 6 months (50%/66%), 9 months (70%/81%). Progression type (radiographic, clinical, both was not associated with OS. CONCLUSION Our novel prognostic scoring system is effective in achieving patient stratification and may guide clinical decision making. Early radiographic progression appears to precede clinical deterioration and may represent true progression in the elderly.