Canadian Society of Surgical Oncology
Canadian Society of Surgical Oncology

 CSSO Operating Grant

The Canadian Society of Surgical Oncology is pleased to announce the establishment of the CSSO Research Fund. This fund is specifically designed to support innovative research by new and emerging surgical oncology researchers, to increase their ability to secure competitive external funding.

One grant valued at $20,000 will be provided per year to a member of the CSSO. Grant proposals covering all aspects of surgical oncology are eligible, including translational research, health services research, quality improvement, clinical trials, clinical epidemiology, surgical education, technology and innovation. Selection will take place through an internal review process, with an objective to provide constructive feedback that can be used to improve future grant applications.

Details concerning eligibility criteria and selection process can be found in the reference document available below. For additional details, please contact ari.meguerditchian@mcgill.ca or info@csso.surgery.

The on-line application form is also available below. The deadline for applications is March 15, 2024.

The Canadian Society of Surgical Oncology (CSSO) Research Grant was made possible through the generous donations of grateful patients, the medical community, and others.  This grant is managed and administered by the Canadian General Surgery Foundation.  For over 40 years, the Foundation has prioritized research, education, and advocacy to ensure that general surgeons in Canada can continue to undertake impactful activities that has optimal patient care at its core. 

To learn more about the work of the Foundation, or to donate the CSSO Research Grant, visit: generalsurgeryfoundation.ca.

Research & Publications

doctors-background-7.jpg

Awareness of Risk Reducing Strategies amongst High Risk Women referred for Breast Cancer Risk Assessment

Basmah Alhassan, MBBS MSc 1, Stephanie M. Wong, MD MPH 2, Marianne Bou Rjeily MD(c) 3, Victor Villareal-Corpuz RN BSc 4, Ipshita Prakash MD MSc 5, Mark Basik MD 6, Jean Francois Boileau MD MSc 7, Michael Pollak MD 8, William D. Foulkes MBBS PhD 9

1 McGill University, Montreal, Canada, King Saud University, Riyadh, Saudi Arabia, 2 Department of Surgery, McGill University Medical School, Montreal, QC, Canada, Stroll Cancer Prevention Centre, Sir Mortimer B. Davis Jewish General Hospital, Montreal, Canada, Department of Oncology, McGill University Medical School, Montreal, QC, Canada, 3 Department of Surgery, McGill University Medical School, Montreal, QC, Canada, 4 Jewish General Hospital Stroll Cancer Prevention Centre, Montreal, QC, Canada, 5 Department of Surgery, McGill University Medical School, Montreal, QC, Canada, Stroll Cancer Prevention Centre, Sir Mortimer B. Davis Jewish General Hospital, Montreal, Canada, Department of Oncology, McGill University Medical School, Montreal, QC, Canada, 6 Department of Surgery, McGill University Medical School, Montreal, QC, Canada, Department of Oncology, McGill University Medical School, Montreal, QC, Canada, 7 Department of Surgery, McGill University Medical School, Montreal, QC, Canada, 8 Stroll Cancer Prevention Centre, Sir Mortimer B. Davis Jewish General Hospital, Montreal, Canada,Department of Oncology, McGill University Medical School, Montreal, QC, Canada, 9 Jewish General Hospital Stroll Cancer Prevention Centre, Montreal, QC, Canada, Department of Oncology, McGill University Medical School, Montreal, QC, Canada, Department of Human Genetics, McGill University Medical School, Montreal, QC, Canada

Introduction: Primary prevention of breast cancer in women at elevated risk includes several strategies such as endocrine prevention and risk-reducing mastectomy. The objective of this study was to evaluate awareness of different preventive strategies across high-risk subgroups.

Methods: Women referred for high risk evaluation at our institution between 2020-2022 completed an initial risk assessment questionnaire that included questions around perceived lifetime risk and consideration of preventive strategies. Wilcoxon Rank Sum and Chi-squared tests were used to compare differences across different high risk subgroups.

Results: Overall, 320 women with a median age of 44 years (IQR, 35-53 years) were referred for high risk evaluation; 120 (37.5%) germline pathogenic variant carriers (GPV), 42 (13.4%) with high-risk lesions (HRL) on breast biopsy, and 158 (49.4%) with strong family history (FH; median Tyrer-Cuzick score 27.7, IQR, 21.9-33.5). Patients from different subgroups reported similarly high levels of perceived lifetime risk (p=0.26) and concern around developing breast cancer (p=0.33). Prior to their risk-assessment, most high risk women reported that they had considered increased screening and surveillance (84.0%) and lifestyle strategies to lower risk (79.6%), while fewer patients had considered risk-reducing surgery (39.3%) or endocrine prevention (27.0%). Risk-reducing mastectomy was most considered by GPV carriers (58.9%) relative to those with HRL (32.4%) or strong FH (25.5%, p<0.001). Based on current guidelines, 123 (38.4%) patients presenting for risk assessment were deemed eligible for endocrine prevention, including 85.7% with HRL and 38.6% with strong FH. Consideration of endocrine prevention prior to the first clinic visit was highest in patients with HRL and significantly lower in those with strong FH (43.2% HRL vs. 32.1% GPV vs. 18.9% FH, p=0.004).

Conclusion: Endocrine prevention is the least considered option for prevention of breast cancer in high risk women, despite eligibility in a significant proportion of those presenting with HRL or strong FH. Further research that focuses on improving awareness around endocrine prevention options and systematically evaluating candidacy is warranted.

Radical chest wall resection and hyperfractionated accelerated radiotherapy (HART) for radiation-associated angiosarcoma of the breast: 15 years of safe and effective treatment
Alexandra Allard-Coutu, Hon BSc, MDCM, MScClin, FRCSC, University of Ottawa, Barbara Heller, Hon BSc, MD, FRSCS, McMaster University

Introduction: Radiation-associated angiosarcomas (RAS) of the breast are rare vascular tumors arising in a previous radiation field for primary breast cancer. A wide range of treatment strategies exist, and while resection remains the definitive treatment, RAS are associated with a high probability of local recurrence and poor overall prognosis. The sarcoma group at the Juravinski Cancer Centre (JCC) previously reported a case series of nine patients treated with radical resection and adjuvant hyperfractionated accelerated radiotherapy (HART). Since 2015, this has become the standard of care at McMaster University, and updated outcomes are reported for a series which now includes 14 patients treated with radical resection and adjuvant HART.

Methods: The JCC pathologic database was reviewed between the year 2006-2022 for patients with RAS. Patients who received radical surgery and immediate HART were eligible. A soft tissue sarcoma approach, rather than a mastectomy approach, was used in the surgical resection, with definitive chest wall soft tissue resection down to the level of the pectoralis muscle with en bloc simple mastectomy, including resection of the previously radiated skin below the clavicle to 2cm inferior to the inframammary crease. Definitive immediate soft tissue coverage was achieved with massive local advancement flaps. Radiotherapy was then delivered to 4500 cGy in 45 fractions three times daily using parallel opposed photon beams and electron patching, or volumetric modulated arc therapy. Primary outcome was recurrence-free survival in months, and records were reviewed for descriptive reports of toxicity. Results were compared to other institutions? experiences.

Results: In our cohort of 14 patients, the recurrence rate was 14.3% (n=2), with a median time to recurrence of 36 months. The overall survival was 77.8% over a median follow-up of 36.7 months (range 3-100 months). One of fourteen patients developed LR and metastasis, four died of other causes, one developed LR and remains disease free after re-resection, and eight are alive with no recurrent disease. Median follow up for patients alive with NED was 25 months (n=9). There were ten reports of mild skin toxicity during treatment. One patient developed chronic wound healing complications which eventually resolved, and one patient developed asymptomatic radiation osteitis of a rib.

Conclusion: Our institutional experience suggests that large normal tissue volume reirradiation with hyperfractionated accelerated radiation therapy is safe and results in improved local recurrence rates when compared to series of patients treated without adjuvant radiation.

The role of anastomotic leak as a potential high-risk criteria in Stage II colon cancer

Ramy Behman, MD PhD 1, Julie Hallet, MD MSc 1, Natalie Coburn, MD MPH 1, Shady Ashamalla, MD MSc 1, Sheron Perera 1, Alyson Mahar, PhD 2, Irene Jeong, MSc 3

1 University of Toronto, 2 University of Manitoba, 3 Institute for Clinical Evaluative Sciences

Introduction: Anastomotic leak (AL) is common following colorectal cancer (CRC) resection. Previous studies have suggested an association between AL and disease recurrence, but results have been mixed. We examined the impact of AL after CRC resection on disease-free and overall-survival (DFS; OS) and compared AL to established high-risk criteria for which adjuvant chemotherapy is indicated in CRC.

Methods: This is a population-based retrospective cohort study of adults with stage I-III CRC undergoing primary resection (2007-2020). Time-to-event analyses compared DFS (accounting for the competing risk of death) and OS between patients with AL and those without. Fine-Gray and Cox models examined the adjusted association between AL and DFS and OS, respectively, with subgroup analyses by disease site (colon/rectum) and stage. Patients were further stratified by presence of established high risk clinical criteria (obstruction and/or perforation) and/or AL to evaluate the association of AL with outcomes, relative to other high-risk criteria.

Results: Of 57,407 patients included (39,907 colon and 17,500 rectum cancers), AL occurred in 5.9%. Median follow-up was 5.2 (IQR: 2.6-8.8) years. After adjustment, AL was associated with significantly lower DFS for colon (sub-hazard ratio, sHR 1.2, 95%CI 1.1-1.3) but not rectal cancer (sHR 1.0, 95%CI 0.9-1.1). The difference in DFS with was greatest among for Stage II colon cancer, with 3-year DFS of 63.5% (95%CI 61.8%-65.1%) with AL vs 71.4% (95%CI 71.0%-71.8%) without AL (p=0.03). AL was associated with lower OS in both colon (HR 1.4, 95%CI 1.3-1.5 and rectal cancer (HR 1.2, 95%CI 1.1-1.3). Tests of the interaction between AL and chemotherapy were not significant for neither DFS nor OS, suggesting that the impact of AL on outcomes was not mediated by its association with receipt of chemotherapy.

Conclusion: AL is independently associated with increased risk of cancer recurrence for patients with colon cancer, but not for those with rectal cancer, and with lower OS for both colon and rectal cancers. The association between AL and DFS is comparable to that of other established high-risk criteria for which adjuvant chemotherapy is indicated. The role of adjuvant chemotherapy for patients with AL after CRC resection warrants further investigation in efforts to improve outcomes

Care trajectory, informational needs and challenges of patients with neuroendocrine tumors: a mixed methods study
Florence Bénard1, Sandrine Huez2, Frédéric Mercier1,3, Marie-Pascale Pomey2, 3, 4

1. Surgical Oncology, Department of Surgery, Centre Hospitalier de l’Université de Montréal (CHUM), 1051 Sanguinet Street, Montreal, QC, H2X 3E4, Canada., 2. Centre d'Excellence pour le Partenariat avec les Patients et le Public, 900 Saint-Denis Street, Montréal, Québec, H2X 0A9, Canada., 3. Centre de recherche du Centre hospitalier de l'Université de Montréal (CRCHUM), 900 Saint-Denis Street, Montréal, Québec, H2X 0A9, Canada., 4. School of Public Health, Université de Montréal, 7101 Parc Avenue, Montréal, Québec, H3N 1X9, Canada.

Introduction: Patients with neuroendocrine tumors (NETs) encounter different challenges, including delayed diagnosis and difficulty finding high-quality NET-specific information. The objective was to document NET patients’ care trajectory and explore their informational needs. The study also aimed to report challenges patients faced throughout the course of their disease, more specifically when trying to obtain information, and identify patient-based solutions.

Method: Thirteen NET patients were recruited and completed a questionnaire to collect sociodemographic and disease-related data. Semi-structured interviews were conducted with each patient alone or with their spouse. Interviews were transcribed verbatim and analysis was conducted using principles of grounded theory. Thirteen patients and four spouses were interviewed. A mean of 26.6 months separated the start of symptoms and the moment of diagnosis. On average, an additional 12.8 months elapsed before referral to a specialized center. 76.5% of patients felt well informed, but only 47.1% believed available information was reliable. Facilitating factors included their relationship with NET specialists, as well as having relatives working in healthcare and a multidisciplinary team, including nurse navigators. Significant challenges included discordances between different sources of information, as well as having limited understanding of medical terms.

In conclusion, there is a need to develop high-quality, vulgarized, accessible sources of information for NET patients. Moreover, rapid referral to NET-specialized centers with access to a multidisciplinary team could ease patients’ care trajectory and facilitate information provision.

Robotic-assisted surgery in general surgery and sub-specialty training: understanding the landscape and perspectives of Canadian program directors

Anna Dare, MBChB PhD 1, Paul Karanicolas, MD PhD 1, Hala Muaddi, MD PhD 1, Richard Walker, MD 1, Simon Laplante 1, Rachel Roke2

1 University of Toronto, 2 Sunnybrook Health Sciences Centre

Introduction: Adoption of robotic-assisted surgery (RAS) is increasing globally, including in general surgery and sub-specialty disciplines. Uptake of the technology in Canada has been slower, and it?s place in Canadian surgical training is debated. The objective of this work was to determine current exposure to RAS during Canadian general surgery and sub-specialty training, and to understand the perspectives of Canadian program directors (PDs) on the integration of RAS in their surgical training programs and curricula.

Methods: A cross-sectional, email-based, national survey of all Canadian general surgery residency and fellowship PDs was conducted in April 2022.

Results: In total, 24/83 PDs responded to the survey: 11/18 (61.1%) were general surgery residency PDs and 13/66 (19.7%) were Fellowship PDs. 91.7% (n=22/24) of respondents reported their trainees have access to a robotic surgical system, however, only 36.3% stated that this is used by trainees in a clinical setting. Only 16.7% reported having a formal RAS curriculum. Most PDs believed that RAS should be part of surgical training in Canada at the fellowship level (72.7%, n=16). Fewer thought it should be part of residency training (36.3%, n=8). A majority (59%) felt that RAS training would be relevant to the practice setting where their graduates will work.

Conclusion: Despite the presence of RAS platforms at most academic centres in Canada, few trainees have hands-on exposure. A gap exists between current RAS training and its anticipated relevance to trainees upon graduation. Focused discussion on the introduction and place of RAS within Canadian surgical training is warranted, considering educational, clinical and health system needs.

Medial location of the primary tumor within the breast: a novel association with local recurrence after skin-sparing mastectomy with immediate reconstruction.
Michael Guo, MD 1, Noelle Davis, MD, FRCSC 1, Sita Ollek, MD FRCSC 2, Leo Chen, MSc 1, Caroline Speers, BA 3, Trevor Hamilton, MD FRCSC 1, Alan Nichol, MD CCFP 3, Caroline Lorisch, MD FRCSC 3, Noelle Davis, MD FRCSC1

1 University of British Columbia, 2 Kelowna General Hospital, 3 BC Cancer

Introduction: Skin-sparing mastectomy (SSM) with immediate breast reconstruction (IBR) is oncologically safe and provides superior cosmetic and psychological outcomes in breast cancer. Medial location of the primary tumor has been suggested to be an independent risk factor for local recurrence (LR) and poor survival. This study aims to elucidate risk factors, including tumor location, for LR in patients undergoing SSM with IBR.

Methods: Patients who underwent treatment with SSM and IBR between 1989-2012 were identified from the BC Cancer Breast Cancer Outcomes Unit database. Univariate and multivariate analyses were used to evaluate predictive factors for LR.

Results: 698 patients were identified. Median length of follow-up was 9.5 [5.3-13.9] years. LR occurred in 14.3% overall, 22% in medial and 12% in lateral tumor groups. On multivariate analysis, medial tumor location (HR 1.98, 95%CI 1.16 ? 3.38, p=0.01) and radiotherapy (HR 1.83, 95%CI 1.05 ? 3.15, p=0.03) predicted higher risk of LR while chemotherapy (HR 0.51, 95%CI 0.27 ? 0.97, p=0.04) and year of treatment (HR 0.92, 95% CI 088 ? 0.97, p<0.001) predicted lower risk of LR.

Conclusion: Medial location of primary breast tumor was significantly associated with LR. We suggest considering this risk factor when selecting the best surgical approach and adjuvant therapy.

Prioritizing melanoma surgeries to prevent wait time delays and upstaging of melanoma during the COVID-19 pandemic

Rebecca Lau*, BMSc and Katherine Aw*, BMSc, Carolyn Nessim, MD, MSc, FRCSC, FACS

Department of Surgery, Faculty of Medicine, University of Ottawa, Ottawa, ON, Canada
*These two authors contributed equally to this work.

Introduction: The COVID-19 pandemic has disrupted the delivery and usage of healthcare services in Canada. For aggressive cancers such as melanoma, prompt diagnosis, staging, and surgical management strongly impact prognosis. Considering the limited resources, emergency closures, and staffing shortages during the pandemic, our institution implemented a dedicated care pathway to prioritize cancer surgeries. We aim to assess whether this strategy was effective at preventing surgical wait time delays and upstaging of melanoma.

Methods: We retrospectively collected data of patients ? 18 years with biopsy-proven primary melanoma who underwent wide local excision between March 1, 2018-February 29, 2020 (pre-COVID) and March 1, 2020-February 28, 2022 (COVID). Patients with distant metastasis, recurrence, and unknown primary were excluded. Wait time from consult to surgery, tumor (T) and nodal (N) stage, and overall stage were collected.??

Results: We included 409 patients [pre-COVID (n = 203) and COVID (n = 206)] with a mean age of 62.2 ± 15.2 years and a male to female ratio of 1.3:1. Average wait time (days) ± SD to surgery was 38.1 ± 23.2 pre-COVID and 40.2 ± 24.0 COVID (p = .482). There were no differences found in T stage (p = .124), N stage (p = .177), or overall melanoma stage (p = .191).

Conclusion: These findings highlight the importance of streamlining melanoma surgery during a pandemic. As we emerge from the pandemic and the need arises to meet surgical backlogs including benign surgery, dedicated cancer surgery should maintain a priority to not negatively affect cancer outcomes.

Population-Level Breast Reconstruction for Immigrant and Long-term Resident Women undergoing Breast Cancer Surgery

Olivia Lovrics MD MSc 1, Elena Parvez MD MSc FRCPC 1 2, David Kirkwood BScH MS 3, Christopher J Coroneos MD MSc FRCSC 4 5, Nicole Hodgson MD MSc FRCSC 1 2, Aristithes Doumouras MD MPH FRCSC 1 3, Jessica Bogach MD MSc FRCSC 1 2

1 Division of General Surgery, McMaster University, 2 Department of Surgery, Juravinski Hospital and Cancer Centre, 3 ICES, Toronto ON, 4 Division of Plastic Surgery, McMaster University, Hamilton, ON, 5 Division of Plastic Surgery, Juravinski Hospital and Cancer Centre, Hamilton, ON

Introduction: Immigrants are susceptible to marginalization, bias, difficulty navigating the health care system. Breast reconstruction surgery after mastectomy for breast cancer is associated with improved quality of life, and access to reconstruction is an important quality of care metric. This study aims to demonstrate differences in breast reconstruction after mastectomy for breast cancer between immigrant women and Canadian long-term residents in Ontario, Canada.

Methods: A retrospective population-based cohort-study using linked provincial administrative databases of patients with Stage I-III breast cancer diagnosed between 2010-2016. Immigration status was obtained from the federal Immigration Refugee and Citizenship Canada database. Variables including, age, stage, co-morbidity, socio-economic factors, cancer histology, and treatments were collected. Data on treatment facility and characteristics were collected. Proportion of immigrant and Canadian long-term resident women undergoing breast reconstruction were compared. Multivariable analysis was performed to adjust for patient, tumour, and treatment characteristics.

Results: 46,930 long-term residents and 7,160 immigrant women. Immigrant women were younger at diagnosis (median 52 vs. 63 years, p<0.01), and more likely to have Stage III disease (16.8% vs. 13.9%, p<0.01). Immigrant women were more likely to be treated at urban, high-volume breast surgery centers with plastic surgeons. 2,196 immigrant women (30.7%) and 13,656 (29.1%) Canadian-long-term residents underwent mastectomy. On univariate analysis, immigrant women were more likely to undergo breast reconstruction surgery overall when compared to Canadian long-term residents (21.4% vs. 18.9%. p<0.01) but were less likely to undergo delayed reconstruction (13.9% vs. 13.3%, p=0.42). Region of origin, but not time in Canada, were significantly associated with reconstruction uptake. On multivariable logistic regression, immigrant women were less likely to undergo reconstruction when adjusting for baseline covariates [OR 0.59 (0.51-0.68)], including patient characteristics, tumour characteristics, and location.

Conclusion: Immigrant women were more likely to undergo breast reconstruction after mastectomy on univariate analysis, but when adjusting for baseline covariates, the inverse relationship was found. This may be that immigrant women are more often treated at urban, high-volume centers with availability of plastic surgeons, and as a result undergo less reconstruction than would be expected when accounting for these factors.

Is Rectosigmoid a distinct primary site of colorectal cancer?

Paul Savage, MD, PhD 1, Carol J. Swallow, MD, PhD 1 2, David P. Cyr, MD, MSc 1, Sameer Shivji, MD 3, James Conner, MD, PhD 3, Richard Kirsch, MBChB, PhD 3, Anand Govindarajan, MD, MSc 1 2, Satheesh Krishna, MD 4

1 Division of General Surgery, Department of Surgery, University of Toronto, Toronto, Ontario, Canada, 2 Department of Surgical Oncology, Mount Sinai Hospital, Toronto, Ontario, Canada, 3 Department of Pathology and Laboratory Medicine, Sinai Health System, Toronto, Ontario, Canada, Richard Kirsch, MBChB, PhD, 4 Medical Imaging, University of Toronto, Toronto, Ontario

Introduction: It is unclear whether primary colorectal cancer (CRC) located at the junction between the sigmoid colon and upper rectum (?rectosigmoid?) should be treated as colon or rectal cancer.

Methods: Consecutive patients who underwent resection of a primary colorectal adenocarcinoma at Mount Sinai Hospital from 2011-2016 were identified from a prospective database (N=800). For tumours distal to the left colon (N=400), pre-treatment CT ± MRI images were reviewed to classify site of primary tumour based on relationship to the sigmoid take-off as defined by D?Souza et al. (Ann Surg 2019). Clinicopathologic features and patient outcomes were compared using Kruskal-Wallis, ?2, Kaplan-Meier and log-rank.

Results: Patients were classified as having primary sigmoid (N=147), rectosigmoid (N=89) or upper rectal (N=70) cancer. The proportion who received neoadjuvant therapy was similar for rectosigmoid vs. upper rectal (53% vs. 54%), and less for sigmoid cancer (17%)(p<0.0001). The early re-operation rate was higher following resection of rectosigmoid vs. sigmoid and upper rectal cancer (8% vs. 3% and 0%, p=0.02). Positive resection margins were more frequent for rectosigmoid vs. upper rectal and sigmoid cancer (16% vs. 4% and 5%, respectively p=0.008). At median follow-up of 4.9 years (IQR 2.7-6.0), local recurrence in patients with rectosigmoid cancer (13%) approximated that for upper rectal cancer, which was higher than sigmoid cancer (HR 7.6, 95%CI 2.2-25.7 rectosigmoid vs. sigmoid, p=0.002).

Conclusion: Rectosigmoid cancers account for >10% of all CRC and carry a notably high risk of positive resection margins and local recurrence. Management of rectosigmoid cancer should be distinct from that of colon cancer.

High Tumour Microenvironment Score is associated with improved overall survival in gastric cancer patients receiving chemotherapy

Daniel Skubleny, MD, PhD , Gina R. Rayat, Msc, PhD, Dan E. Schiller, MD, MSc, Jennifer Spratlin, MD, Sunita Ghosh, PhD

University of Alberta

Introduction: Gastric cancer remains an aggressive disease and is the third most common cause of cancer death worldwide. Molecular heterogeneity in gastric cancer is associated with variable treatment responses to cytotoxic chemotherapy. We evaluated whether complex molecular classification systems in gastric cancer could inform subgroup treatment effects related to chemotherapy.

Methods: Molecular subtypes of the Cancer Genome Atlas (TCGA), Asian Cancer Research Group (ACRG), and Tumour Microenvironment (TME) Score were learned from whole transcriptome data using artificial intelligence and applied to a public cohort of 2,202 gastric cancer patients. Bias between gene expression measurement technologies was mitigated using feature-specific quantile normalization. We used propensity score matching to identify comparable patients treated with and without chemotherapy. Matching covariates included molecular subtype scores and clinical characteristics. We assessed survival outcomes using multivariable Cox proportional hazards models.

Results: Our matched cohort included 237 patients treated with chemotherapy and 158 patients who did not receive chemotherapy. The absolute standardized mean difference between covariates was <0.1. Increasing TME high score was associated with improved survival in patients receiving chemotherapy compared to those that did not (Hazard Ratio 0.47 [95% CI 0.29, 0.74], Interaction p=0.04). TCGA and ACRG molecular subtypes as well as tumour stage and location provided no subgroup treatment effect related to chemotherapy.

Conclusion: In a propensity score matched cohort, we identify that gastric cancer patients who receive cytotoxic chemotherapy achieve greater overall survival as the TME score increases. These data could improve patient selection for chemotherapy but must be further validated in a prospective study.

Surgical decision making in young breast cancer patients: Impact of pre-treatment surgical preference congruence with final surgical treatment on psychosocial health

Melissa L. Wood, MD, BSc, MSc, FRCSC 1, Dr. May Lynn Quan, MD, BSc, MSc, FRCSC 1, Dr. Kelly Metcalfe, RN, PhD 2, Dr. Yuan Xu, MD, PhD 1, Dr. Susan Isherwood, PhD 1, Yue (Flora) Yang, MDSA 1, Xing Wang, MSc

1 University of Calgary, 2 University of Toronto

Introduction: Type of breast surgery in young women is known to impact quality of life. We sought to determine if congruence between pre-treatment surgical preference and final surgery had an impact on psychosocial health.

Methods: Women < 41 with breast cancer in the RUBY study. Patient demographics, pre-consult surgical preference, surgery type, psychosocial outcomes, and congruence were collected. Outcomes were decision regret, anxiety, PHQ-9, and BreastQ. Univariate data analyses were performed.

Results: Among 1000 women, median age was 37 years; 39.2% had breast conserving surgery and 60.8% mastectomy. In total, 326 women expressed congruence, 206 non-congruence, while 468 had no preference. Reasons for mastectomy differed significantly; ?what I wanted? (congruent 87.4%, non-congruent 55.7%, no preference 70.9%; p < 0.001) and ?surgeon recommended? (congruent 39.4%, non-congruent 74.5%, no preference 60.3%; p < 0.001). At baseline, global anxiety was significantly higher (p = 0.02) in the non-congruent compared to the congruent/no preference groups. Depression, decision regret, and overall anxiety were not significantly different post-surgery. When stratified by surgery type, using the BreastQ assessment, post-treatment breast satisfaction (p = 0.03), and psychosocial wellbeing (p = 0.05), were significantly poorer in the non-congruent group.

Conclusion: Non-congruent status of preop preference to actual surgery is significantly associated with higher baseline global anxiety and poorer psychosocial wellbeing. Awareness of these psychosocial implications may provide an opportunity to develop targeted support.

  • CSSO-01

    Incidence, Timing and Outcomes of Venous Thromboembolism in Patients Undergoing Surgery for Esophagogastric Cancer: A Population-Based Cohort Study

    Hanna Nader, Williams Erin, Kong Weidong, Fundytus Adam, Booth Christopher, Patel Sunil, Caycedo-Marulanda Antonio, Chung Wiley, Nanji Sulaiman, Merchant Shaila, Queen's University

    Background: Abdominal surgery and chemotherapy are risk factors for venous thromboembolism (VTE) in patients with cancer, but their contribution in patients with esophagogastric cancer is unclear. We quantified VTE risk, identified risk factors for VTE, and determined the association between VTE and survival in patients undergoing surgery for esophagogastric cancer.

    Methods: A population-based retrospective cohort study was conducted using linked administrative databases. We used the Ontario Cancer Registry to identify patients with esophageal or gastric cancer between January 2007 and December 2016 who underwent surgery. First VTE event was captured at clinically relevant timepoints 180 days before and after surgery. Logistic regression was used to identify factors associated with VTE with odds ratios (OR) and 95% confidence intervals (CI) reported. Cox proportional hazards regression models were used to estimate associations between covariates and overall survival (OS) and cancer-specific survival (CSS).

    Results: Of the 4,894 patients who had esophagectomy or gastrectomy, 8% (n=383/4,894) had a VTE. VTE risk was 2.5% (n=123/4,894) 180 days before surgery, 2.8% (n=138/4,894) within 30 days of surgery and 2.5% (n=122/4,894) from 31- ≤180 days after surgery. Of the patients with VTE within 30 days of surgery, 34% (n=47/138) were diagnosed after hospital discharge. Pre-operative chemotherapy was associated with VTE 180 days before surgery (OR 3.84, 95% CI 2.41, 6.11). Increased hospital length of stay (LOS) was associated with VTE 30 days after surgery (OR 1.08, 95% CI 1.02, 1.14, per week). In adjusted models VTE was associated with inferior OS (HR 1.36, 95% CI 1.13, 1.63) and CSS (HR 1.42, 95% CI 1.16, 1.75).

    Conclusion: Highest VTE risk is within 30 days of surgery with one third diagnosed after hospital discharge. Longer hospital LOS and pre-operative chemotherapy are associated with increased VTE risk. VTE is an independent risk factor for inferior survival in patients with esophagogastric cancer.

    CSSO-04

    Omission of Axillary Staging and Survival in Elderly Women with Early Stage Breast Cancer: A Population-Based Cohort Study

    Castelo Matthew, Hansen Bettina, Paszat Lawrence, Baxter Nancy, Scheer Adena , University of Toronto

    Introduction: Surgical axillary staging in women ≥ 70 years with early stage breast cancer is controversial. Older randomized evidence has not shown axillary staging improves survival, but recent observational studies have been mixed and widespread de-implementation of the practice has not occurred. The aim of this study was to determine if axillary staging is associated with survival in elderly women with breast cancer.

    Methods: This was a population-based cohort study using the SEER registry. Women ≥ 70 years diagnosed with T1-T2 invasive breast cancer from 2005 to 2015 were included. Overlap propensity score weighting was used to adjust for confounders. Overall survival (OS) was determined and hazard ratios (HRs) reported with 95% confidence intervals (CIs). Breast cancer-specific survival (BCSS) was determined using competing risks analysis, and subdistribution hazard ratios (sdHRs) reported. Additional adjustment was performed for receipt of chemotherapy and radiotherapy.

    Results: 144,329 elderly women were included, of whom 22,621 (15.7%) did not undergo axillary staging. After overlap propensity score weighting, baseline characteristics were well balanced between the two groups. Women who did not undergo axillary staging were significantly less likely to receive chemotherapy (adjusted RR 0.58, 95% CI 0.54 - 0.62) or radiotherapy (adjusted RR 0.53, 95% CI 0.52 - 0.54), and had significantly worse OS (adjusted HR 1.22, 95% CI 1.19 - 1.25), and breast cancer-specific survival (adjusted sdHR 1.14, 95% CI 1.08 - 1.21) compared to those that had staging. Subgroup analyses restricted to women with ER+/HER2- tumours showed similar findings to the main analysis (adjusted sdHR 1.17, 95% CI 1.05-1.31).

    Conclusion: These findings suggest elderly women with early-stage breast cancer who do not undergo axillary staging experience worse outcomes. Reasons for this disparity may be multi-factorial and require further investigation.

    CSSO-05

    Patients' Experiences Receiving Cancer Surgery during the COVID-19 Pandemic: A Qualitative Study

    Pook Makena1, Najafi Ghezeljeh Tahereh 2, Lapointe-Gagner Maxime3, Nguyen-Powanda Philip3, Elhaj Hiba4, Rajabiyazdi Fateme 5, Lee Lawrence4, Feldman Liane S. 4, Fiore Jr. Julio6

    1 Division of Experimental Surgery, McGill University, Montreal, QC, Canada, 2 School of Nursing and Midwifery, Iran University of Medical Sciences, Tehran, Iran., 3 Division of Experimental Surgery, McGill University, Montreal, QC, Canada., 4 Steinberg-Bernstein Centre for Minimally Invasive Surgery and Innovation, McGill University Health Centre, Montreal, QC, Canada., 5 Department of Systems and Computer Engineering, Carleton University, Ottawa, ON, Canada., 6 Centre for Outcomes Research and Evaluation, Research Institute of the McGill University Health Centre, Montreal, QC, Canada.

    INTRODUCTION: In response to COVID-19, Quebec repurposed surgical care infrastructure and delayed many elective cancer surgeries. However, postponing cancer surgery is known to cause anxiety and distress. A qualitative study was conducted to understand patients' experiences receiving surgical cancer treatment during the COVID-19 pandemic.

    METHODS: Patients who underwent general surgery for cancer at the McGill University Health Centre between March 2020 and January 2021 were invited to one-to-one interviews. Patients were purposefully selected for maximum variation using quota sampling (i.e., targeting delay status, pandemic phase, cancer site, and clinical/demographic characteristics) until interviews produced no new information (i.e., thematic saturation). Interviews were conducted using a semi-structured guide, audio-recorded, transcribed verbatim, and analyzed independently by two researchers. Data were managed using MAXQDA2020 and analyzed according to inductive thematic analysis.

    RESULTS: Interviews were conducted with 20 patients [mean age 64; male (n=10); cancer sites: breast (n=8), skin (n=4), hepato-pancreato-biliary (n=4), colorectal (n=2), and gastro-esophageal (n=2)]. Surgery was delayed for 14 patients: 8 by the hospital, 4 by the patient, and 2 due to a positive COVID-19 test. Thematic analysis revealed that patients considered their susceptibility to infection, hospital safety measures, and burden on healthcare resources when determining willingness to undergo surgery. Patients weighed these risks against the urgency of their health condition and recommendations of their provider. Changes to the hospital environment (e.g., COVID-19 preventative measures) and deviations from expected treatment (e.g., alternative treatments, remote consultations, rescheduled care) caused diverse psychological responses, ranging from increased satisfaction to severe distress. Patients employed coping strategies (e.g., reframing care interruptions, communicating with clinicians, information seeking) to mitigate distress.

    CONCLUSION: In summary, changes in care during the pandemic elicited diverse psychological responses from patients undergoing cancer surgery. Patient coping was facilitated by open, consistent communication with clinicians, emphasizing the importance of patient-centered discussions regarding surgical delays within and beyond the pandemic.

    CSSO-08

    Quality of Narrative Central and Lateral Neck Dissection Reports for Thyroid Cancer Treatment Suggests Need for a National Standardized Synoptic Operative Template

    Watanabe Akie1, Prisman Eitan 1, Mitmaker Elliot2, Walker Ross3, Wu Jonn1, Nguyen Anne1, Wiseman Sam1

    1 University of British Columbia, 2 McGill University, 3 Queen's University

    Introduction: Consistent documentation of anatomic structures in central (CND) and/or lateral neck dissections (LND) is important for effective communication between multidisciplinary teams. This study aimed to investigate the current completeness of CND and LND narrative operative reports.

    Methods: Twenty-nine surgeons from 6 provinces who treat malignant thyroid disease provided de-identified CND and LND narrative reports. Important report elements were identified based on recommended items from prior literature and summarized using descriptive statistics for both CND and LND (stratified by dissection level).

    Results: Amongst 53 CND reports, 66% and 17% documented level VI and VII dissections, respectively. 25% did not indicate the level(s) of dissection. Other than the recurrent laryngeal nerve(s) (96%), status of critical structures including the carotid artery(s) (43%), superior laryngeal nerve(s) (15%), and innominate artery(s) (9%) were insufficiently reported. Amongst 23 LND reports, 9%, 83%, 83%, 87%, and 61% documented level I, II, III, IV, and V dissections, respectively. Status of the sternocleidomastoid muscle (91%) and internal jugular vein (91%) were frequently reported across all dissection levels, while only 39% recorded the presence or absence of a chyle duct leak. For level I dissections (N=2), the status of the submandibular gland(s), lingual nerve(s), and hypoglossal nerve(s) were reported 100% of the time. Similarly, the integrity of the spinal accessory nerve(s) (86%) was frequently reported for level V dissections (N=14). In contrast, important structures such as the vagus nerve(s) (50%), cervical rootlets (27%), and occipital artery(s) (5%) for level II/III dissections (N=22) and the thoracic duct (20%) for level IV dissections (N=20) were inadequately reported.

    Conclusion: Current narrative operative reporting fails to consistently document the status of important structures dissected in the central and lateral necks. Development of an accepted standardized national synoptic operative template may enhance reporting completeness and facilitate improved quality of patient care across multidisciplinary teams.

    CSSO-09

    Transoral Endoscopic Thyroidectomy Vestibular Approach (TOETVA): Indications and Technique

    Saravana-Bawan Bianka1, Hong Dennis2, Gupta Michael2, Pasternak Jesse1

    1 University Health Network, University of Toronto, 2 St. Joseph's Healthcare, McMaster University

    INTRODUCTION: Thyroid surgery has progressed from large collar incisions to more cosmetically sensitive dimensions. Nevertheless, scarring remains a prominent concern among patients specifically from communities where neck scar has substantial stigma. To address these concerns, alternatives to traditional approach were explored, such as trans-axillary and facelift approaches.

    METHODS: These approaches, however, require robotic instrumentation, considerable extra training, and cutaneous incisions. Comparatively, transoral endoscopic thyroidectomy vestibular approach (TOETVA) uses common laparoscopic equipment and has minimal learning curve for endocrine surgeons trained in thyroid surgery and laparoscopy.

    RESULTS: In a recent Johns Hopkins review, learning curve for TOETVA was noted to be 7 to 11 cases. Once learning curve has plateaued, TOETVA operative times are not significantly longer than that of traditional thyroidectomy, demonstrating an average of 78 compared to 64 minutes for thyroid lobectomy and 135 compared to 103 minutes for total thyroidectomy. This approach is comparatively easy for thyroid surgeons to learn as it employs the traditional thyroidectomy planes. Importantly, there are no increased risks of complications in comparison to traditional thyroidectomy as seen in the largest series. Paratracheal tissue, including parathyroid glands, are easily observed and the recurrent laryngeal nerves (RLNs) are visualized in a plane which represents a more favorable angle with magnified view than that of traditional open approach. Patient selection is key as benign nodules less than 6cm with thyroid lobe no larger than 10cm are recommended for TOETVA. As expertise and familiarity increases, these limits may be pushed.

    CONCLUSION: Originally performed for benign disease alone, TOETVA is now also performed for small (less than 2cm), well differentiated thyroid cancers. In addition to these criteria, patients with a short mental distance and good neck extension should be considered. The only contra-indication to TOETVA is the inability to tolerate surgery or general anesthetic. This video demonstrates TOETVA performed at our centre.

    CSSO-11

    Molecular Landscape of Early-Stage Breast Cancer with Nodal Metastasis

    Brackstone Muriel, Ghasemi Farhad

    Western University

    Introduction: The presence of axillary lymph-node metastasis is an important prognostic factor in breast cancer, and serves as the basis of important treatment decisions such as neoadjuvant or adjuvant systemic therapy. Clinical evaluation of axilla lymphadenopathy is inaccurate, leading to sentinel lymph-node biopsies (SLNBs) to stage the disease. SLNB is associated with morbidity for the patients, excludes the possibility of future SLNB, and requires health-care resources. As such, a better understanding of molecular processes leading to axillary metastasis in breast cancer is important and can aid us in pre-operative prediction of nodal involvement.

    Methods: A multi-platform comparison of early-stage breast tumours (≤ 5cm) in patients undergoing SLNB was compared using data from The Cancer Genome Atlas (TCGA) project. The comparison between 250 node-negative and 162 node-positive early-stage breast tumours revealed 766 statistically significant differentially expressed mRNAs and 40 differentially expressed miRNAs.

    Results: Distinct heterogeneity existed between the four molecular subtypes (Luminal A, Luminal B, Basal, Her2) of breast cancer. Only 33.2% of the discovered differentially expressed mRNAs were either over-expressed or under-expressed consistently across all subtypes with nodal involvement. Pathway enrichment analysis highlighted several pathways including immune-response and chromatin assembly. There were no statistically significant differences in single nucleotide variations, copy number alterations or protein expression between node-negative and node-positive patients.

    Conclusions: The potential molecular signatures identified in this study may prove valuable in the development of predictive models of axillary involvement, and highlight the importance of a subtype-specific approach to breast cancer.

    CSSO-14

    Association between Patient-Reported Symptoms and Health Care Resource Utilization: A First Step to Develop Patient-Centred Value Measures in Cancer Care

    Hirpara Dhruvin 1, Eskander Antoine 2, Coburn Natalie2, Sutradhar Rinku3, Chan Wing 3, Hallet Julie 2

    1 University of Toronto, 2 Sunnybrook Hospital, 3 ICES

    Introduction: Value of care is defined as patient-relevant outcomes achieved per dollar spent. Patient Reported Outcomes (PROs) offer a unique lens into cancer care; their relationship with cost and resource utilization, however, is yet to be defined. We examined the association between PROs and health resource utilization (HRU) in the year after cancer diagnosis, with a view to develop PRO-based measures of Value.

    Methods: We conducted a population-based cohort study of adults with cancer (2008-2019). The exposure was symptom burden measured using Edmonton Symptom Assessment System (ESAS) scores. The outcome was total healthcare cost within 30-days of ESAS reporting, as a metric for HRU. HRU was further stratified into cancer-directed therapies (i.e., chemotherapy and radiation) or ancillary services including emergency department visits, complex continuing care, homecare, and inpatient mental health. Linear regression models with log-transformed costs examined the association between ESAS scores and outcomes adjusting for potential confounders.

    Results: 1,728,025 ESAS surveys from 285,924 patients were analyzed. Gastrointestinal, breast and CNS cancers were the most resource intensive cancers with median 30-day costs (CAD) after ESAS of $85,000, $81,000, and $78,000, respectively. Each 10-point increase in total ESAS score (0 to 90) was associated with a 9.4% decrease in the cost of cancer-directed therapies. Conversely, each 10-point increase was associated with an 18% increase in costs of ancillary care. The association between symptom burden and HRU is depicted in Figure 1.

    Conclusions: High symptom burden was associated with decreased use of cancer-directed therapy but increased use of ancillary care, indicating interruptions to oncologic treatment. Proactive symptom management may mitigate unnecessary HRU and facilitate cancer-directed therapies. Future work will further explore the relationship between PROs and HRU/costs to develop a PRO-based measure of Value in cancer care.

    CSSO-12

    Beta Testing of a Risk-Stratified Patient Decision Aid to Facilitate Shared Decision Making for Postoperative Extended Thromboprophylaxis in Patients Undergoing Major Abdominal Surgery for Cancer

    Ivankovic Victoria1, Delisle Megan1, Stacey Dawn2, Abou-Khalil Jad2, Balaa Fady3, Bertens Kimberly2, Dingley Brittney3, Martel Guillaume2, McAlpine Kristen4, Nessim Carolyn2, Tadros Shaheer3, Carrier Marc2, Auer Rebecca2

    1 The University of Ottawa, 2 Ottawa Hospital Research Institute, 3 University of Ottawa, 4 University of Toronto

    Background: We previously developed a novel patient decision aid (PtDA) to facilitate shared decision-making between patients and clinicians when deciding whether to use extended-duration thromboprophylaxis or not for four weeks after major abdominopelvic surgery for cancer. Our PtDA was found to be acceptable with patients and clinicians. The objective of this study was to build on our previous work by evaluating the effectiveness of our PtDA.

    Methods: Patients undergoing major abdominal surgery for cancer at an academic centre were enrolled in this pre- post-test study. Institutional ethics approval was obtained. All outcomes were measured using previously psychometrically validated surveys. The primary outcome was change in decisional conflict. Secondary outcomes included readiness to make a decision, confidence in decision making, and change in patient knowledge of the health care decision. A sample size calculation determined a total of 17 patients were required to demonstrate the PtDA meaningfully reduced decisional conflict using a paired t-test.

    Results: A total of 17 patients were recruited. The median age was 68 years old (range 28-82) and the majority of patients were male (13/17, 76.5%). Based on the Caprini Score, 1 patient was low risk, 6 were moderate risk, 5 were high risk, and 5 were very high risk of developing a venous thromboembolism. The median pre-PtDA decisional conflict score was 2.4, compared to a post-PtDA score of 1.3 (p<0.01). The median score for confidence in decision-making was 86.4, corresponding to high confidence. Median knowledge scores increased from 50% to 75%. Median score for readiness to make a decision following the PtDA was 90, indicating a high perceived level of preparedness to make a decision.

    Conclusions: The PtDA significantly reduced decisional conflict and was effective at improving the parameters of patients' decision-making abilities. Patients demonstrate high confidence for decision-making, and indicate they are prepared to decide after using the PtDA.

    CSSO-16

    Why Do Patients with Non-metastatic Primary Retroperitoneal Sarcoma Not Undergo Resection?

    Ng Deanna1, Acidi Belkacem 2, Johnston Wendy3, Callegaro Dario4, Brar Savtaj3, Gladdy Rebecca3, Chung Peter 5, Catton Charles5, Khalili Korosh6, Honore Charles2, Swallow Carol3

    1 University of Toronto, 2 Gustave Roussy, 3 Mount Sinai Hospital, 4 Istituto Nazionale dei Tumori, 5 Princess Margaret Cancer Centre, 6 University Health Network

    INTRODUCTION: Resection is the mainstay of management of primary RPS, but an unknown proportion of patients do not undergo resection even though the tumour is localized. Very few centres systematically collect data regarding RPS patients who do not come to resection. We investigated the incidence of and underlying reasons for non-resection, using prospectively maintained data from two high volume sarcoma centres.

    METHODS: Consecutive patients who presented with primary RPS and no distant metastases on staging imaging were included (n=276; 2012-2017).

    RESULTS: Of these, 188 (69%) underwent resection, while 88 (31%) did not. Patients who did not have resection were older (Table), and approximately half had significant comorbidities (n=46) and/or poor performance status (n=41). Interestingly, the median maximum tumour size was smaller in the non-resected cohort. The 88 patients who did not undergo resection were divided into 3 groups. Group A (n=23) were patients who were deemed technically unresectable due to extensive involvement of the SMA/SMV, portal vein, aorta, spinal canal or mediastinum. Group B (n=40) comprised patients who received either no (n=29) or brief (n=11) active treatment, with no short-term (3 months) progression of disease. Group C (n=25) were patients who progressed on what was planned to be cytoreducing preoperative treatment (chemotherapy and/or XRT) (n=25). For the entire cohort of 88 non-resected patients, median OS was 13 months and 3yr OS was 35% (95% CI 25-48%) (Kaplan Meier). Patients in Group A (technically unresectable) and Group B (no progression) had similar OS, but Group C (progression on treatment) showed a rapid decline in OS (Figure, p=0.007). Nearly one-third of patients who presented with non-metastatic primary RPS did not ultimately undergo resection.

    CONCLUSION: Progression of disease on planned preoperative treatment was associated with very limited survival, revealing adverse biology. The ability to predict early progression would facilitate adaptive response to guide innovative therapeutic approaches.

    CSSO-17

    Loss of FAM46C Expression Predicts Inferior Post-resection Survival and Induces Ion Channelopathy in Gastric Adenocarcinoma

    Luu Shelly1, Fu Ning2, Kazazian Karineh1, Pacholczyk Karina3, Ng Deanna1, Swett-Cosentino Jossie4, Savage Paul1, Shibahara Yukiko5, Kalimuthu Sangeetha1, Espin-Garcia Osvaldo1, Conner James1, Yeung Jonathan1, Darling Gail6, Swallow Carol1

    1 University of Toronto, 2 University of Ottawa, 3 Lunenfeld-Tanenbaum Research Institute, 4 Brockville General Hospital, 5 Kitasato University School of Medicine, 6 Dalhousie University

    INTRODUCTION: More precise delineation of recurrence risk and pattern would facilitate personalized use of adjunctive therapies in patients with gastric adenocarcinoma (GCa). Our aim is to discover high fidelity markers of risk, and novel therapeutic targets.

    METHODS: Tumour (T) and paired normal mucosa (NM) samples were microdissected from curatively resected GCa specimens from 158 patients (2001-2017). RNA was extracted and differential gene expression correlated with patient outcome. Disease-specific survival (DSS) was estimated by the Kaplan-Meier method and hazard ratios estimated with Cox Regression.

    RESULTS: Median age of the study cohort was 70 years, with a median post-resection follow-up time of 31 months (IQR 12-73) and 3-year DSS of 66%. We explored potential markers of recurrence risk by performing a directed screen of 55 members of the oncogene Plk4 interactome. This revealed reduced expression of the nucleotidyl transferase FAM46C in tumour tissue in 94% of patients.

    CONCLUSION: Retention of FAM46C expression (T/NM≥median, n=79) was associated with superior 3-year DSS (75% vs 57% in T/NM

    CSSO-18

    Liver-Directed Therapy of Neuroendocrine Liver Metastases

    Meloche-Dumas Léamarie1, Mercier Frédéric2, Barabash Victoria3, Law Calvin4, Coburn Natalie4, Singh Simron4, Myrehaug Sten4, Chan Wing5, Hallet Julie4

    1Université de Montréal, 2Centre hospitalier de l'Université de Montréal (CHUM), 3Sunnybrook Research Institute, 4Odette Cancer Centre/Sunnybrook Health Sciences Centre, 5Cancer Research Program, Institute of Clinical Evaluative Sciences (ICES)

    INTRODUCTION: The optimal therapy sequencing for metastatic neuroendocrine tumors (NETs) remains undefined. Recent advances in systemic therapies may have changed approaches. Better understanding in patterns of care is necessary to assess and design treatment strategies.

    METHODS: We examined the use of factors associated with liver-directed therapy over time. We conducted a population-based study of metastatic NETs over 2000-2019. Outcomes were use of liver-directed therapy, sub-divided into liver resection and embolization. Bi-yearly incidence rate of use in eligible patients (alive and no prior liver-directed therapy) was assessed. Multivariable Poisson models examined factors associated with use of liver-directed therapies.

    RESULTS: Of 5,159 metastatic NETs, 922 patients (16.7%) received liver-directed therapy (461 embolizations, 329 resections, 132 dual therapy) at median of 35 days (IQR:0-490) after metastatic diagnosis. Incident use of liver embolization increased after 2013 to reach 72% in 2018-19. Incident use of liver resection followed a similar trajectory up to 94% in 2018-19 (Figure). Gastro-entero-pancreatic primary NET (relative risk - RR 5.69; 95%CI 3.76-8.60), female sex (RR 1.25 95%CI 1.05-1.48), year of diagnosis (RR 1.32; 95%CI 1.04-1.68 for 2007-2015), and lower socioeconomic status (RR 0.93, 95%CI 0.87-0.98 by incremental material deprivation quintile) were independently associated with liver resection. Gastro-entero-pancreatic primary NET (RR 2.8, 95%CI 2.2-3.7), socioeconomic status (RR 0.94, 95%CI 0.89-0.99 by quintile) and year of diagnosis (RR 0.71, 95%CI 0.59-0.85 for 2007-15 and RR 0.61, 95%CI 0.50-0.75 2016-2020) were independently associated with risk of liver embolization.

    CONCLUSION: Receipt of liver-directed therapies for metastatic NETs has increased over time in unadjusted analysis. However, there was lower risk of liver embolization in most recent time periods, but higher risk of resection. Socio-economic status represented an independent factor for lower likelihood of liver-directed therapies. Further characterization of timing and outcomes of liver-directed therapy, with an equity lens, is warranted to define the optimal sequencing.

    CSSO-20

    Molli® for Excision of Non-palpable Breast Lesions: A Case Series

    Eom Ashley, Muhn Narry, Heller Barbara, Lovrics Peter

    McMaster University

    INTRODUCTION: Nonpalpable breast lesions require intraoperative localization. Magnetic Occult Lesion Localization Instrument (MOLLI®) is a novel localization technique that employs ferromagnetic marker technology, detected through a hand-held probe for intra-operative lesion localization, without the limitations of radioactive seeds or wire localizations.

    METHODS: We examined the clinical outcomes of MOLLI®-guided localization of non-palpable breast lesions at a single institution. A consecutive sample of 30 patients with non-palpable breast lesions underwent lumpectomy with or without sentinel lymph node biopsy. MOLLI® markers were placed pre-operatively by a breast radiologist under sonographic or stereotactic mammogram guidance. The hand-held probe was used to localize the marker signal intra-operatively. Complete excision of the marker and lesion were confirmed with specimen radiography. Patient demographics, operative, pathological data were collected retrospectively from electronic medical records.

    RESULTS: 7 of 30 patients had multifocal disease requiring multiple markers. One patient had bilateral lesions. One patient failed marker placement as it did not deploy within a cystic lesion. Thus, a total of 29 patients and their 36 lesions were analyzed and summarized in Table 1. One pathologic margin was involved by DCIS, but clear of invasive carcinoma. Marker migration did not occur. Post-operatively, two patients developed cellulitis and one patient developed hematoma.

    CONCLUSION: Our study is the second series to demonstrate the MOLLI® marker to be an effective alternative to radioactive seeds in the localization of non-palpable breast lesions. All markers were successfully identified and yielded surgical specimen with margins clear of invasive carcinoma. The markers were placed up to 47 days pre-operatively without complications, allowing for flexibility in scheduling without the resources required for radiation safety. One marker failed to deploy within a cystic lesion; the deployment mechanism was subsequently altered to correct this. The novel MOLLI® technology guides surgery with dynamic audible signals, visual guidance, and precise distance measurements, enabling efficient localization.

    CSSO-24

    Absence of Benefit of Routine Surveillance in Very Low and Low Risk Gastric Gastrointestinal Stromal Tumors

    Schmitz Erika1, Apte Sameer2, Nessim Carolyn 3

    1 University of Ottawa Faculty of Medicine, 2 University of Ottawa, 3 The Ottawa Hospital, Department of Surgical Oncology

    INTRODUCTION: Gastric gastrointestinal stromal tumors (GIST) are mesenchymal neoplasms with heterogenous malignant behavior. Adjuvant therapy and routine surveillance is guided by the risk of recurrence, which is largely determined by tumor location, mitotic rate, size and intra-operative tumor rupture. Recurrence after surgical resection of very low and low risk gastric GISTs is exceedingly rare. Despite this, the National Comprehensive Cancer Network suggest abdominopelvic cross-sectional imaging every 3-6 months for 3-5 years then annually, while the European Society for Medical Oncology state that routine follow-up may not be warranted in very low risk GIST and the benefit of routine follow-up is not known. Consequently, the aim of this study is to characterize the patterns of recurrence amongst the low and very low risk gastric GISTs, and determine the value of surveillance at our center.

    METHODS: Adult patients with gastric GIST who underwent surgical resection at a single tertiary care center between 2010-2020 were evaluated. Demographics, clinical presentation, radiologic and endoscopic findings, pathological results and surveillance data were collected and analysed.

    RESULTS: 139 patients underwent resection of a gastric GIST and were eligible for inclusion. According to the National Institute of Health modified classification system, 8.6% (n=12) were considered very low risk, 37.4% (n=52) low risk, 36% (n=50) intermediate risk and 18% (n=25) high risk. We observed one recurrence in the intermediate risk group at 4 years (2%) and two within the high-risk groups at 2 and 3 years (8%), all of which were non-perforated, asymptomatic and detected on routine surveillance imaging. Amongst the very low risk group, three were discharged to their family physician for surveillance, and the remaining 77.8% (n=7) underwent surveillance with cross sectional abdominal imaging and 33.3% (n=3) with additional chest imaging, respectively. Amongst the low-risk group, ten patients were discharged to their family physician for surveillance, and the remainder 57.1% (n=24) underwent surveillance with cross sectional abdominal imaging and 30.9% (n=13) with additional chest imaging, respectively. Nine patients of the low-risk group underwent endoscopic surveillance. After a median of 37.1 and 34.3 months surveillance within the very low and low risk groups, we observed no recurrences and detected no additional malignancies.

    CONCLUSION: There were no recurrences of very low risk and low risk gastric GIST after surgical resection in this single site population. While considering cost-effectiveness and in the absence of randomized control trials, this evidence may support that routine radiologic and endoscopic surveillance may not be warranted amongst these subgroups.

    CSSO-26

    Active Surveillance for DCIS of the Breast: Qualitative Interviews with Patients and Physicians

    Newman-Bremang Jieun, Look Hong Nicole, Gagliardi Anna, Nyhof Bryanna

    University of Toronto

    INTRODUCTION: Multiple international trials are currently investigating the safety of active surveillance (AS) for low-risk DCIS, where surgical excision is offered only in the case of progression to invasive cancer. The objective of this study was to explore views on acceptability of AS for low-risk DCIS from DCIS survivors and physicians.

    METHODS: Women with a history of DCIS were recruited using purposive sampling and interviewed in five focus groups. Concurrently, physicians specializing in breast cancer care were recruited via purposive and snowball sampling and interviewed via semi-structured telephone interviews. All interviews were conducted using interview guides that were created and pilot tested as part of a larger study on patient centered care on DCIS. The responses were recorded and transcribed verbatim. Data was analyzed iteratively with qualitative descriptive analysis and constant comparative analysis was used to extract dominant themes.

    RESULTS: We interviewed a total of 35 women in five focus groups held in 5 provinces in Canada (British Columbia, Alberta, Saskatchewan, Ontario, Nova Scotia), as well as 40 physicians from general surgery, surgical oncology, medical oncology, radiation oncology and radiology. The majority of DCIS survivors and physicians were hesitant towards active surveillance, and risk of missed invasive disease was the top concern by both patients and physicians. Both groups recognized that it is not yet the standard of care, and addressed the need for tools to predict the risk of progression prior to recommending AS.

    CONCLUSION: The DCIS survivors and physicians were overall aligned in their hesitancy for AS in its current state. If AS were to be proven to be safe, these important considerations should be included for an effective implementation of AS for DCIS.

Oncology Journals