Princess Margaret Cancer Centre Research Articles

Prognostic Role of Multiparameter Flow Cytometry-Based Measurable Residual Disease Assessment in Acute Myeloid Leukemia Patients with FMS3-like Tyrosine Kinase-3 Internal Tandem Duplication (FLT3-ITD)

Introduction Measurable residual disease (MRD) monitoring is predictive in acute myeloid leukemia (AML). Assessment of FMS3-like tyrosine kinase-3 in-frame internal tandem duplications (FLT3-ITD) is usually performed at diagnosis by polymerase chain reaction (PCR). Due to its relative lower sensitivity of 2%, FLT3-ITD PCR is not routinely used during response assessment. Next-generation sequencing is similarly limited by technical difficulties in capturing tandem duplications using the short base pair-based sequencing method. Long base pair-based sequencing has been reported, but its use is limited by financial restrictions. Multiparameter flow cytometry (MFC) can be a useful tool for MRD monitoring in this AML subtype until such time that molecular techniques for detecting FLT3-ITD MRD are optimized. Patients and methods The study evaluated the outcomes of FLT3-ITD mutated AML patients diagnosed and treated from 2018 to 2022 at Princess Margaret Cancer Centre. We compared outcomes according to MFC-MRD post-induction and FLT3-ITD allele frequency (AF) status at diagnosis. MRD cut-off was 0.1%. Data were locked as of June 30, 2023. Clinical outcomes evaluated include overall survival (OS) and relapse-free survival (RFS). The cumulative incidence of relapse (CIR) and non-relapse mortality (NRM) were calculated considering competing risk. The Kaplan-Meier method using a log-rank test and a multivariate Cox proportional hazard model was used for analyses, while the Gray test and Fine-Grey model were used for uni- and multivariate analysis for CIR and NRM. Results A total of 111 patients with a mean age of 63.5 years were included, of whom 90 received treatment. Secondary AML accounted for 12.7% of patients. Risk stratification according to European LeukemiaNet (ELN) 2022 was favorable in 2 (1.8%), intermediate in 67 (60.4%), and adverse in 42 patients (37.8%). Nucleophosmin 1 (NPM1) co-mutation was observed in 55 patients (49.5%). Seventy-nine patients (87.8%) could be assessed for overall response, including 69 (76.7%) who achieved complete remission (CR) or CR with incomplete count recovery (CRi). Of these, 54 achieved first CR/CRi (CR1) with 1 induction cycle. MFC-MRD data were available in 61 patients, of whom 44 (72.1%) were MRD negative, while 17 (27.9%) were MRD positive. With a median follow-up of 437 days, 50 patients (45%) were still alive. Median OS and RFS were 3.42 years and 1.05 years, respectively. Among patients who achieved CR1, post-induction MFC-MRD positivity correlated with an inferior OS (HR 2.35 [1.06-5.26], p=0.037) and a trend for a shorter RFS (HR 2.08 [0.99-4.35], p=0.052). We examined the impact of FLT3-ITD AF at diagnosis on long-term outcomes. By applying a binary recursive partitioning method, the cut-off of FLT3-ITD AF with the best risk stratification power for RFS, was defined at 54.6%. The group with a higher FLT3-ITD AF showed inferior OS (HR 1.86 [1.01-3.44], p=0.047) and RFS (HR 1.91 [1.09-3.33], p=0.023). Taking together FLT3-ITD AF at diagnosis and MFC-MRD status at CR1, patients with low FLT3-ITD AF and negative MRD had the highest OS rate of 86.2% at 12 months (p=0.023), and the highest RFS at 72.4% (p=0.096), while the corresponding values for low FLT3-ITD AF/positive MRD patients were 87.5% and 50%, respectively. In contrast, those with high FLT3-ITD AF/negative MRD had an OS of 72.7% and a RFS of 45.5%, while those with high FLT3-ITD AF/positive MRD showed the lowest OS (16.7%) and the shortest RFS (16.7%). There was no statistical difference in CIR and NRM among groups. Multivariate analysis with stepwise selection was performed, considering age at diagnosis, ELN 2022 risk, MFC-MRD at CR1, FLT3-ITD AF at diagnosis, cytogenetics, and NPM1 co-mutation. Predictive factors for OS were MFC-MRD post-induction (HR 2.49 [1.11-5.61], p=0.027) and FLT3-ITD AF (HR 2.29 [1.03-5.11], p=0.043). For RFS, age at diagnosis (HR 1.03 [1.00-1.06], p=0.038) and FLT3-ITD AF (HR 2.40 [1.15-5.01], p=0.019) were predictive, while MFC-MRD was not significant (HR 1.59 [0.72-3.53], p=0.25). Conclusion Our data demonstrate that MFC-based MRD assessment is feasible in AML with FLT3-ITD. Although better outcomes are expected in patients with a lower FLT3-ITD AF, patients who failed to achieve MRD negativity at CR1 showed inferior outcomes. The presence of both poor risk factors, a high AF of FLT3-ITD at diagnosis and MRD positivity at CR1, correlated with the worst treatment outcomes.

Clinical, Cytogenetic and Immunophenotypic Spectra of Post-Transplant Relapse Following Allogeneic Hematopoietic Stem Cell Transplantation for Myelodysplastic Syndrome

Introduction Allogeneic hematopoietic stem cell transplantation (HCT) is a curative therapy used to treat myelodysplastic syndrome (MDS). Relapse post-HCT confers poor outcomes and subsequent therapy is not standardized. Like acute myeloid leukemia (AML), several mechanisms are known to be involved in post-HCT relapse including clonal evolution in favor of chemotherapy-resistant clones and tumor immune escape. Systemic reviews are scarce in the MDS post-HCT relapse population. Accordingly, the present study aimed to investigate clinical, cytogenetic, and immunophenotypic (IP) spectra for MDS post-HCT relapse patients (pts), along with subsequent therapies, and clinical outcomes following post-HCT relapse. Patients and methods A retrospective study was conducted and found that 231 MDS pts received HCT between April 2010 and September 2022 at the Princess Margaret Cancer Centre (PMCC). The primary endpoint was overall survival (OS) following post-HCT relapse. Secondary endpoints explored time to relapse (TTR), relapse presentation (i.e., MDS or AML), clonal evolution (i.e., changes in cytogenetics or IP profiles), and therapeutic strategies delivered after relapse. Results Of the 231 MDS pts treated with HCT, 52 (23%) relapsed in a median of 4.5 months (range 0.9-59.8 months). Twenty-five (48%) pts relapsed with MDS (i.e., blast <20%), while 27 (52%) relapsed with AML (i.e., blast ≥20%) in the marrow. Fifteen (65%) pts showed cytogenetic changes at the time of post-HCT relapse. Eleven (48%) of which presented with more aggressive findings (i.e., additional structural abnormalities and monosomal karyotype). Ten of 17 (59%) pts presented with changes in their IP profile at post-HCT relapse, 7 (70%) of which relapsed with AML. Combined these findings give evidence of clonal evolution in favor of more aggressive malignant clones that are often responsible for post-HCT relapse. Treatment with hypomethylating agents (HMA) alone or in combination after post-HCT relapse increased OS to 5.6 months in comparison to those not receiving HMA of whom OS was 1.4 months (p=0.0003). Stratified by disease presentation cytogenetic clonal evolution was seen in 6/12 (50%) of the post-HCT MDS relapses, and 9/11 (82%) of the post-HCT AML relapses. Cytogenetic abnormalities at diagnosis and post-HCT relapse, respectively included in descending order: structural abnormalities (n=14/23, 61%; n=20/23, 87%), complex karyotype (n=12/23, 52%; 17/23, 74%), -5/del(5q) (n=11/23, 48%; n=10/23, 43%), -7/del(7q) (n=8/23, 35%; n=11/23, 48%), monosomy karyotype (n=8/23, 35%; n=11/23, 48%), -17 (n=3/23, 13%; n=5/23, 22%), der(5)/t(5;17) (n=2/23, 9%; n=5/23, 22%), -20 (n=2/23, 9%; n=3/23, 13%), and inv(3) (n=1/23, 4%, n=2/23, 9%). Seventeen pts had paired IP profiles available both prior to HCT and at post-HCT relapse, of which 10 (59%) had profile changes of at least 1 antigen. Seven of 10 (70%) pts that relapsed with AML post-HCT had IP expression changes, compared to 3/7 (42.9%) pts that relapsed with MDS post-HCT. Summarized IP findings include reduced expression of CD7 (n=2/6, 33%), altered expression of CD34 (n=3/15, 20%), altered expression of CD117 (n=3/14, 21%), and reduced expression of HLA DR (n=4/15, 27%). The presence of IP expression changes at post-HCT relapse suggests evidence of clonal evolution or immune escape as the mechanism behind the relapsed disease. Overall survival duration was 3.97 months; the 12-month OS rate was 20.0% (10.0-32.5%). Post-HCT relapse treatments included hypomethylating agents (HMA; n=13/46, 28%), HMA with donor lymphocyte infusion (DLI; n= 7/46, 15%), HMA with Venetoclax (VEN; n=7/46, 15%), chemotherapy (n=9/46, 19%), best supportive care (8/46, 17%) and second HCT (n=2/46, 4%). Patients that received HMA with or without DLI or VEN had a longer survival duration of 5.6 months (95% CI [3.4-13.2 months]) compared to those not receiving HMA 1.4 months (95% CI [0.6-0.39 months, p=0.0003]). The 12-month survival rate was also significantly in favor of pts that received HMA following post-HCT relapse (35.3% [95% CI 17.2-54.0%] vs 4.3% [95% CI 0.3-18.0%]). Conclusion Patients with MDS undergoing HCT may relapse with either MDS or AML. The relapsed disease often shows evidence of clonal evolution in its immunophenotype and/or cytogenetic profile toward a more aggressive disease. Approaches to post-HCT relapse remain heterogenous but HMA is beneficial.

The Presence of Somatic Mutation in Donor Does Not Affect Long-Term Outcomes in 339 Patients Following Allogeneic Hematopoietic Stem Cell Transplantation Using Uniform Fludarabine/Busulfan Myeloablative Conditioning and Cyclosporine/Methotrexate and ATG Gvhd Prophylaxis

Introduction: Clonal hematopoiesis with indeterminate potential (CHIP) constitutes a part of the biological aging process and comprises the acquisition of somatic mutations in hematopoietic stem cells (HSCs). In the context of allogeneic hematopoietic stem cell transplantation (HCT), the potential transfer of CHIP from donor to recipient may raise concerns on HCT-related complications. There are contradictory data on donor CHIP's impact on long-term transplant outcomes using very diverse conditioning regimen as well as various graft-versus-host disease (GVHD) prophylaxis. This study aims to evaluate the impact of donor CHIP on allogeneic HCT outcomes in a cohort of the patients transplanted using a uniform, myeloablative intensity conditioning of Fludarabine-Busulfan (4 days) as well as Cyclosporine/Methotrexate with 4.5 mg/kg rabbit antithymocyte globulin (ATG) over 3 days for GVHD prophylaxis. Patients and methods: Peripheral blood samples were obtained from recipients and their corresponding donors who underwent allogeneic HCT at Tom Baker Cancer Centre, Calgary, Canada. Samples were archived from 339 patients and 255 donors prior to HCT, and additional patient sampling was done on day 28 (D28), day 56 (D56), or day 84 post allogeneic HCT (D84). As a result, paired recipient-donor samples were available for 255 HCTs. This study was approved by the institutional review boards of Princess Margaret Cancer Centre, Toronto, and Tom Baker Cancer Centre, Calgary. For the detection of CHIP, we have applied a bar-coded error-corrected sequencing protocol, a single-molecule molecular inversion probe capture protocol (NAR Genom Bioinform, 2022), which was performed at Wiesmann Institute of Science (Rehovot, Israel). We designed probes targeting 34 genes covering CHIP mutations along with other AML related mutations. Bar-coded NGS library was generated and processed. All statistical analyses were performed using R statistical software 3.5.0 and EZR version 1.41. Results: Sixty-seven clonal mutations in 56 recipients (18.2%) and 22 clonal mutations in 18 donors (7.1%) were detected prior to HCT. For recipients, TET2, JAK2, RUNX1, PPM1D, and DNMT3A were detected in at least 2 patients or more. For donors, the most frequently mutated gene was TET2 (3.1%), and the seconds were DNMT3A and PTPN11 (1.2% for both). Baseline characteristics of 255 patient-donor pairs did not show any difference between D-CHIP and no D-CHIP groups except donor age. Donor CHIP status was distributed evenly throughout most of the demographics of recipients, type of disease, pre-transplant status, and transplant procedure. Most of the pts received peripheral blood stem cells (97%). With a median follow up duration 6.02 years, the 5-year OS rates tend to be higher in no D-CHIP group than D-CHIP group (63.1% vs 42.8%) with a HR of 1.73 (95% CI [0.90-3.35]), although it was not statistically significant (p=0.096). The cumulative incidence of relapse for both groups were not different in 5 years (28.1% vs. 28.4%, p=0.849). The cumulative incidence of NRM within 5 years in D-CHIP group was slightly higher than no D-CHIP group (28.4% vs 17.5%), but it was not statistically significant (p=0.250). Neutrophil engraftment by day 28 was 99.1% vs. 100%, without statistical difference between the two groups (p=0.785). The cumulative incidence of grade II-IV acute GVHD within day+100 seemed to be higher in D-CHIP group compared to no D-CHIP group, but it was not statistically significant (24.8% vs. 34.5%, p=0.148; Figure A). In terms of chronic GVHD, cumulative incidence of moderate to severe GVHD was significantly higher in 5 years after transplant (29.4% vs. 58.8%; HR 2.55 [1.331-4.96], p=0.006; Figure B). Even when we set the cut off level of donor CHIP positivity up to >1% or >10%, the association between donor CHIP and risk of chronic GVHD remained significant. The impact of donor CHIP on the cumulative incidence of moderate to severe GVHD was confirmed in multivariate analysis (HR 2.3 [1.3-4.1], p=0.003). Other factors associated with moderate to severe chronic GVHD were previous history of grade II-IV acute GVHD (p=0.008) and donor type (p=0.054). Conclusion: Under clinical setting of uniform conditioning regimen and GVHD prophylaxis incorporating ATG (4.5mg/kg), donor CHIP was found to be associated with higher risk of moderate to severe chronic GVHD, while it does not affect other allogeneic HCT outcomes.

Characterization of Clonal Hematopoietic of Indeterminate Potential (CHIP) Mutations in an Imid-Naïve Multiple Myeloma (MM) Autologous Stem Cell Transplant (ASCT) Population: First Results from a Pre-Transplant Time Point in a Prospective, Longitudinal Study

Background: Clonal Hematopoiesis of Indeterminate Potential (CHIP) is defined by the acquisition of ≥ 1 somatic mutations in the blood of healthy adults, but can also be detected at higher prevalence in patients with malignancies, particularly after DNA damaging therapy. The presence of CHIP mutations in MM confers a reduced overall survival (Mouhieddine, JCO 2020) but high quality, prospective evaluation of CHIP mutation evolution over the transplant sequence, including the impact of high-dose melphalan and immunomodulatory (IMiD)-based therapy, is largely under-studied. Serial assessment of CHIP over the transplant course has been reported from the coMMpass dataset, demonstrating a 4-fold increase in mutation prevalence from 5.8% at diagnosis to 25% after 3.1 years (Mouhieddine, Blood, 2021). In this data set, most patients were exposed to prior IMiD-containing at some point during therapy. Whether the survival benefit reported with IMiD-based maintenance therapy post-transplant is associated with CHIP mutation modulation is unclear. We are therefore investigating the longitudinal evolution of CHIP in a cohort of transplant-eligible, largely IMiD-naïve patients (pts) over the pre- and post-ASCT course in an ongoing study. Here, we present preliminary findings from the pre-transplant time-point after non-IMiD-containing induction but before high-dose melphalan and ASCT. Methods: The ongoing Princess Margaret Cancer Center ARCH 001 trial is a prospective, longitudinal study evaluating evolution of CHIP in a transplant-eligible MM population. Mutation testing is performed using the highly sensitive single-molecule molecular inversion probe (SmMIP) next-generation sequencing (NGS) technique (Abelson, Bioinformatics, 2022). After sequencing with a minimum sequencing depth of 4000x, the list of mutation calls produced by smMIP-tools is subject to various filters to reduce the likelihood of false positives. Synonymous mutations, as well as any variants falling within introns or splice regions are removed. The minor allele frequency of the variant (if available) is required to be above 0.1% so as to exclude mutations that are common SNPs. VAF frequency threshold of 1-30 % is set for calling somatic mutations, with a view to minimize false positives and germ-line mutations. Multiple time points for testing pertinent to therapy include: within 1 month pre-transplant (after induction and stem cell collection), 3 months post-transplant, and two subsequent samples 12 and 24 months post-sample 2. Here we report preliminary results of CHIP mutation testing in the first 66 pts at the pre-transplant time point. Results: Patient, disease and treatment characteristics for all pts and their stratification by the presence of CHIP mutations are outlined in Table 1. The total cohort is typical for a transplant-eligible population with median age 65 years, male predominance (M 59%/F 41%), 60% IgG subtype, 23% high-risk FISH cytogenetics. Most (90%) received a non-IMiD containing induction regimen CyBorD; only 6% received an IMiD pre-ASCT. CHIP mutations were identified in 28/66 pts, for a mutation prevalence rate of 43%; 6 patients (9%) carried ≥ 1 mutation. The median age at transplant as well as other baseline demographics did not differ between those with and without mutations (Table 1). Mutation profiles are shown in Figure 1, and are consistent with those described in the literature with the most commonly mutated genes being DNMT3a, TET2, ASXL1 and PPM1D, and the mutation prevalence rate highest for DNMT3a at 21%. 6 pts (9%) had ≥ 1 mutation (range 2-3) identified, with 3 of 6 involving the DNMT3a gene. Other genes involved in patients with multiple mutations were TET2 (n=3), ASXL1 (n=2), PPM1D (n=2), SRSF2 (n=1), SF3B1 (n=1). Mean allele frequency was 4%, ranging from 1-19. Conclusion: Our results using deep SmMIP sequencing show a mutation prevalence rate of 43% in a cohort of transplant-eligible, largely IMiD-naive MM pts studied after induction. These rates are higher than those generally described in the literature, which may relate to the high sensitivity of the assay used, variation in filtering between institutions, and the deliberate use of a relatively low VAF threshold with the intent to capture clinically significant variants for longitudinal tracking during the course of this ongoing study. Further studies during lenalidomide maintenance will be forthcoming.

DISP-05. UNDERSTANDING THE INDIVIDUAL CHALLENGES AND EXPERIENCES OF YOUNG ADULTS WITH A DIAGNOSIS OF GLIOBLASTOMA AND HIGH-GRADE ASTROCYTOMA

Abstract High-grade gliomas are highly aggressive and fatal primary malignant brain tumors in adults with ineffective existing treatment methods. Older adults are greatly impacted by high-grade gliomas as they are more common among this population. Young adults (individuals aged 18-39) are uniquely impacted by high-grade glioma. However, there are few resources and information available to address these unique needs. The purpose of this study is to explore the lived experience of this population to inform the development of resources to help address their specific needs. A qualitative descriptive design drawing on thematic analysis was chosen to conduct and analyze interviews with young adult high-grade glioma patients at the Princess Margaret Cancer Center. Interviews focused on patient experiences as a high-grade glioma patient at Princess Margaret Cancer Center, the challenges they face and changes they would like implemented. A coding framework was created to determine emerging themes using qualitative data analysis software NVivo 10. To date, 5 interviews (1 men; 4 women; age range 19-37) have been completed. Interviews described the challenges living with high-grade glioma as a young adult and the changes in resources and support services patients would like implemented. Experiences highlighted 1) difficulty accepting this change in life and disruptions brought upon life goals and plans at the young adult age; 2) an overall lack of young adult high-grade glioma specific education, advocacy and support resources 3) A call for more community building and connecting so patients do not feel alone through this diagnosis. These patient interviews highlight a connection between the young adult age and high-grade glioma diagnosis and the challenges experienced and the services required. These results will serve as the foundation to advocate for the creation of tools that cater to patient needs of young adult high-grade glioma patients.

Frailty Impact on Outcomes of Patients Undergoing Chimeric Antigen Receptor T-Cell (CAR T) Therapy at Princess Margaret Cancer Centre: A Prospective Pilot Study

Introduction Princess Margaret Cancer Centre is 1 of 3 adult centres in Ontario, Canada providing Chimeric Antigen Receptor T-Cell (CAR T) therapy as a standard of care treatment for adult patients with relapsed and refractory B-cell lymphomas. There is no upper age limit for treatment consideration, and frailty may be an important factor in assessing fitness for treatment. This study aims to determine if frailty assessments pre-CAR T can predict those at higher risk for acute toxicities, PFS, and OS, as well as evaluate changes in frailty over time. Methods We performed a cohort study of consecutive patients with lymphoma undergoing CAR T cell therapy at our institution, from April 2021 to date. Frailty was evaluated using the Clinical Frailty Scale (CFS), Grip Strength, Gait Speed, Mini-Cog, Edmonton Symptom Assessment System (ESAS) and Patient Health Questionnaire (PHQ-2 and PHQ-9) at 5 time points: baseline (clearance visit), and 1, 3, 6 and 12 months post-CAR T. At baseline, additional assessments were completed to further characterize frailty including the Vulnerable Elders Survey (VES-13), Hematopoietic Cell Transplantation-specific Comorbidity Index (HCT-CI) and Cumulative Illness Rating Scale (CIRS). We present data on patients enrolled to date. Results Fifty-two patients have had their data analyzed thus far. Mean age is 57.9 ± 12.7 years and 54% are male. 50% have de novo diffuse large B-cell lymphoma, 24% transformation from follicular lymphoma,14% HGBCL, 6% PMBCL and 6% FL. Most patients (83%) received axi-cel. Median follow-up is 3.80 months (IQR 0.56-24.7 mo). All 52 patients completed assessments at baseline, 44 completed 1 month, 25 completed 3 month, 12 completed 6 month and 11 completed 12 month assessments. Median scores (and IQR) for the HCT-CI, CIRS, VES-13 and CFS at baseline were 1.5 (0-5), 3.0 (0-13), 1.0 (0-7), and 3.0 (1-7) respectively. There were clinically significant changes observed in CFS over time ( p=<0.001), with mean scores of 3.3 at baseline, 3.7 at 1 month, 2.8 at 3 months and 2.1 at 12 months. There were no significant changes ( p=0.051; p=0.078) between timepoints for either of the physical assessments of grip strength or gait speed. Sixteen patients (31%) experienced immune effector cell neurotoxicity syndrome (ICANS) (4% Grade 4, 4% Grade 3, 8% Grade 2, 15% Grade 1) and forty-nine patients (94%) experienced cytokine release syndrome (CRS) (2% Grade 3, 65% Grade 2, 27% Grade 1), during the 30 days following cell re-infusion. Seven patients (13%) were admitted to the ICU, with a median in-hospital length of stay for all 52 patients of 12.5 days (IQR 7.0, 99.0). None of the variables tested (age, sex, bridging therapy, LDH, CRP and frailty assessments) were significantly associated with the development or grade of CRS or ICANS. There have been 21 progression events and 18 deaths (6 without progression); 9 pts remain alive post-progression (Figure 1). On univariable analysis of baseline data, ECOG performance status, LDH level, and the VES-13 score were predictive of PFS, while ECOG, LDH, CRP, VES-13 score, CFS and the time to walk 4m were predictive of OS. The HCT-CI and CIRS score were not significantly associated with outcomes. On multivariable analysis of PFS, baseline VES-13 score (HR 1.17, p=0.04) and LDH (HR 1.002, p=0.05) remained significant. Multivariable analysis for OS was not performed due to insufficient number of events. Analyses incorporating repeated measures were performed, and on univariable analysis, the CFS, 4m walk test (s), LDH, VES-13 were significantly associated with PFS and the same variables and CRP were associated with OS (Table 1). On multivariable repeated measurements analyses, only the CFS was still associated with PFS, and CFS and LDH with OS (Table 1). Conclusions Conducting serial frailty assessments in patients undergoing CAR T therapy is feasible. The only longitudinal measurement of frailty found to have clinically significant changes between timepoints was the CFS, suggesting an element of reversible functional impairment related to patients' lymphoma. Within the limits of our sample size, baseline measures of frailty were not predictive of CRS or ICANS; the relationship of CFS change over time with PFS and OS may be indicative of lymphoma response. Enrollment is ongoing and data on larger number of patients with longer term follow-up are needed.

Quality of colorectal and anal cancer care delivered during the COVID pandemic.

432 Background: Reports suggest that there have been significant impacts to the provision of cancer care with delays, interruptions and cancellations across all treatment modalities as a result of the COVID pandemic which have implications for the quality of care. We evaluated the impact of the early phase of the pandemic relative to the same period in the year prior using a panel of measures spanning the six domains of quality; 8 measures focused on pandemic-specific care modifications, and 16 established measures. Methods: The cohort consisted of all new patient consultations between 02/19 to 12/19 (comparator) or 02/20 to 12/20 (COVID) at Princess Margaret Cancer Centre (PM) in Toronto, Canada, who were >18 years of age and newly diagnosed with colon, rectal or anal cancer. Chart abstraction data was linked to Census and the Ontario Marginalization Index datasets to derive additional population-weighted sociodemographic variables. A summary quality score across established measures was computed and a benchmark was set using the pared mean approach representing the top 10% of performers. Associations between achieving the quality benchmark and patient characteristics were evaluated using a multivariable logistic regression model. Results: Relative to the year prior, there was a 12.2% reduction new patient consultations (294 vs 335). Significant findings for individual indicators are summarized below. Relative to English-first language patients, those whose first language was not English were 7.4 times less likely to achieve the benchmark (OR 0.13; 95% CI 0.01-0.61). Compared to those with stage I disease, patients with stage IV disease at diagnosis were 6.5 times less likely to achieve the benchmark (OR: 0.15; 95% CI 0.02-0.78). Conclusions: While overall quality of care was poorer during the early phase of the pandemic, there was a reduction in the proportion of patients treated with systemic therapy within 30 days of death. This likely reflects efforts to prioritize fitter patients with curative disease for treatment and reduce avoidable healthcare utilization. Future work should focus on understanding the downstream impacts of these differences in care quality on clinical outcomes and optimizing preparedness for future disasters.[Table: see text]

Multi-task machine learning of the electronic medical record to predict future symptoms among patients with cancer.

587 Background: Cancer and its treatments often cause symptoms. Automated warning systems could mitigate symptoms by alerting healthcare teams and enabling personalized preventative interventions. We developed a general-purpose longitudinal system for predicting symptomatic deterioration among outpatients undergoing intravenous systemic anti-cancer therapy. Methods: Patients treated for aerodigestive cancers at the Princess Margaret Cancer Centre were randomly divided into development and testing cohorts. For each treatment, machine learning was applied to preceding electronic medical record (EMR) data to predict patient-reported symptom deterioration, defined as at least a four point worsening on the Edmonton Symptom Assessment Scale. Features included diagnostic and treatment characteristics, laboratory tests, and patient-reported symptoms. Single-task (e.g., LASSO and XGboost) and multi-task (e.g., temporal CNNs, LSTM and Transformer) models were trained, tuned, and evaluated based on discrimination, calibration, and net benefit. Results: The cohort consisted of 3,998 patients who underwent 45,904 treatment sessions, with data across 400 features. Among these patients, 1,547 (38.6%) were female; median age was 64.0 (interquartile range 13.0). The most common diagnoses were lung (1,505, 37.6%), head and neck (696, 17.4%), and pancreatic cancers (685, 17.1%). The best model, a multi-task transformer, predicted symptom deterioration with an AUROC range of 0.732-0.822, marking a 1.4-6.2% improvement over the best single-task model. At a 10% alert rate, treatments associated with alerts would be enriched 4-13 fold for symptom deterioration (P<0.001). The system was calibrated and would provide a net benefit across a wide range of threshold probabilities in decision curve analysis. Conclusions: Longitudinal general-purpose multi-task machine learning systems trained using EMR data can accurately predict a wide range of symptoms. Based on these results, automated warning systems for symptoms should be implemented and evaluated in real-time clinical practice to guide preventative interventions.[Table: see text]

2 The Contribution of Brain Metastases to Neurocognitive Functioning in Patients with Advanced Metastatic Cancer

Objective:Neurocognitive decline is prevalent in patients with metastatic cancers, attributed to various disease, treatment, and individual factors. Whether the presence of brain metastases (BrMets) contributes to neurocognitive decline is unclear. Aims of this study are to examine neurocognitive performance in BrMets patients and compare findings to patients with advanced metastatic cancer without BrMets. Here, we present baseline findings from an ongoing, prospective longitudinal study.Participants and Methods:English-speaking adults with advanced metastatic cancers were recruited from the brain metastases and lung clinics at the Princess Margaret Cancer Centre. Participants completed standardized tests (WTAR, HVLT-R, BVMT-R, COWAT, Trailmaking test, WAIS-IV Digit Span) and questionnaires (FACT-Cog v3, EORTC-QLQ C30 and BN20, PROMIS Depression(8a) and Anxiety(6a)) prior to cranial radiotherapy for those who required it. Test scores were converted to z-scores based on published normative data and averaged to create a composite neurocognitive performance score and domain scores for memory, attention/working memory, processing speed and executive function. Neurocognitive impairment was defined according to International Cancer and Cognition Task Force criteria. Univariate and multivariate regressions were used to identify individual, disease and treatment variables that predict cognitive performance.Results:76 patients (mean (SD) age: 63.2 (11.7) years; 53% male) with BrMets were included. 61% experienced neurocognitive impairment overall; impairment rates varied across domains (38% memory, 39% executive functioning, 13% attention/working memory, 8% processing speed). BrMets quantity, volume, and location were not associated with neurocognitive performance. Better performance status (ECOG; ß[95%CI];-0.38[-0.70,-0.05], p=0.021), higher premorbid IQ (0.34[0.10,0.58], p=0.005) and greater cognitive concerns (0.02[-3.9e-04,0.04], p=0.051) were associated with better neurocognitive performance in univariate analyses. Only premorbid IQ (0.37[0.14,0.60], p=0.003) and cognitive concerns (0.02[0.0004, 0.03], p=0.05) remained significant in multivariate analysis. We also recruited 31 patients with metastatic non-small cell lung cancer (mNSCLC) with no known BrMets (age: 67.5 (8.3); 32% male) and compared them to the subgroup of BrMets patients in our sample with mNSCLC (N=32; age: 67.8 (11.7); 53% male). We found no differences in impairment rates (BrMets/non-BrMets: Cognitive Composite, 59%/55%; Memory, 31%/32%; Executive Functioning, 35%/29%; Attention/working memory, 16%/13%; Processing speed, 7%/6%; Wilcoxon rank-sum test, all p-value’s > 0.5). The presence or absence of BrMets did not predict neurocognitive performance. Among patients with mNSCLC, higher education (0.11[0.03,0.18], p=0.004) and premorbid IQ (0.36[0.12,0.61], p=0.003), fewer days since primary diagnosis (0.00290[-0.0052,-0.0005], p=0.015) fewer pack-years smoking history (0.01[0.02,-0.001], p=0.027) and greater cognitive concerns (0.02[7e-5,0.04], p=0.045) were associated with better neurocognitive performance in univariate analyses; only premorbid IQ (0.26[0.02,0.51], p=0.04) and cognitive concerns (0.02[0.01,0.04], p=0.02) remained significant in multivariate analysis.Conclusions:Cognitive impairment is prevalent in patients with advanced metastatic cancers, particularly affecting memory and executive functioning. However, 39% of patients in our sample were not impaired in any domain. We found no associations between the presence of BrMets and neurocognitive function in patients with advanced cancers prior to cranial radiation. Premorbid IQ, a proxy for cognitive reserve, was associated with cognitive outcomes in our sample. Our longitudinal study will allow us to identify risk and resilience factors associated with neurocognitive changes in patients with metastatic cancers to better inform therapeutic interventions in this population.

Development and implementation of a patient reported experience measurement program to transform cancer care.

343 Background: Improving the patient experience is central to providing high-quality, value-based care. Patient reported experience measures (PREMs) are tools to evaluate and improve quality of care. Your Voice Matters (YVM) is a validated survey created by Cancer Care Ontario that captures patient experience in outpatient settings at Regional Cancer Centres across Ontario, Canada. While most hospitals actively collect patient experience data, few have moved to action feedback to improve outcomes. Princess Margaret Cancer Centre (PM) implemented a systematic approach to PREMs evaluation to advance operational excellence in an outpatient setting. Methods: PM distributes the YVM survey monthly to patients with outpatient visits via email using a Redcap survey platform (93% of patient emails are on file). In addition to the standardized YVM questions, the survey captures respondent comments in an open text field asking about one thing PM can improve and comments are thematically coded and trended monthly. Patients reporting a very poor experience with care (defined as an overall experience of 1/5) who wish to connect with PM are contacted to provide additional information. Key performance indicators (KPIs) were established and visualized in run charts in a Power BI dashboard. KPIs were disseminated to staff and leadership to identify opportunities for improvement. Results: Starting in 2022, approximately 10,600 surveys were disseminated monthly to patients and 1,100 surveys returned (10% response rate) with 10-15 patients contacted each month regarding their poor experience. Long wait times was the top concern across three main outpatient clinics; in the Gynecology clinics, in March 2022, only 45% (35/77) of respondents rated wait times as very good (5/5) and 32% (11/34) of comments were wait times complaints. A comprehensive root cause analysis found that the visit booking system, distribution of clinics across the weekdays, and supports in clinic were key contributors to long wait times. In response, Gynecology clinic resources were reorganized to launch a new Physician Assistant clinic on the underutilized Friday afternoon clinic timeslot, a new streamlined visit process was established, and the booking system was optimized. Within six months, there was a 39% improvement in wait times on weekdays, a 60% increase in the number of patients waiting less than 30 minutes, and a 37% increase in visit capacity on underutilized Fridays. The corresponding patient experience data over the same timeframe improved seen as an increase from 45% to 52% of patients who rated wait times to see the provider as 5/5. Conclusions: A programmatic approach to PREMs evaluation and response can drive improvement in practice. Facilitators include a data-driven methodologic approach to quality improvement, leadership engagement at all levels, and elevation of frontline staff as local experts and change agents.

Addressing challenges in telephone triage for outpatient oncology care: A data-centred digital routing solution.

591 Background: As cancer care shifts to a predominately outpatient model, there is growing recognition that optimizing communication between patients and their care providers between scheduled visits can improve efficiency and patient outcomes. Princess Margaret Cancer Centre (PM), which sees over 80,000 patients annually, operates with many parallel telephone triage lines as the primary means of patients-provider communication. When a patient needs to contact their care team, they are instructed to leave a voicemail and wait for a response. In a high-volume setting this model presents many challenges to the patient experience, provider efficiency, and patient safety. The purpose of this pilot project was to design, implement, and assess the feasibility of a digital channel for patients to reach their care team. Methods: We used a human-centred design approach to develop a new digital triage service. First, we analyzed manually collected triage call data over a one-year period to understand the most common patient queries of the telephone triage service. Through co-design, a new digital-triage service was conceptualized and prototyped. Subsequently, we tested a data-centred digital routing algorithm for patient concerns and symptoms. To evaluate implementation and assess feasibility, we collected usage data, satisfaction surveys, and conducted staff shadowing and interviews. We used PDSA cycles to refine user experience and clinical workflows. Results: Patient adoption of digital triage increased continuously over 9 months, with volumes surpassing voicemail after four months: over the last four months digital triage has encompassed on average 63% of the total request volumes. The response rate to patient surveys was 21%. Patients indicated being very satisfied or satisfied (74.4%), neutral (6.7%), and dissatisfied or very dissatisfied (18.9%). Interviews with leading triage staff suggest the digital channel improves triage efficiency and redirecting/escalation by providing structured and complete patient queries. Conclusions: The pilot highlights the feasibility and acceptability of a digital-triage service in a comprehensive cancer care setting. By leveraging digital tools and real-time data, PM addressed the challenges of telephone triage, enabling a more efficient process to respond to patients’ needs, while demonstrating high adoption and patient satisfaction. Subsequent work will expand the use of digital triage to all disease sites across the cancer centre, and create foundational workflows and technology for proactive care models, such as Remote Patient Monitoring.

Dental Extractions Before Radiation Therapy and the Risk of Osteoradionecrosis in Patients With Head and Neck Cancer

Patients with head and neck cancer undergo extraction of teeth with poor prognoses to minimize post-radiation therapy (RT) extractions, which are known to cause osteoradionecrosis (ORN). However, many patients are required to start RT before the extraction sites are completely healed. The role of pre-RT extractions in the development of ORN has been disputed in literature. To determine whether the timing of pre-RT dental extractions is associated with ORN development in patients with head and neck cancer. This retrospective cohort study was conducted at a single institution (Princess Margaret Cancer Centre, Toronto, Canada) between January 1, 2011, and January 1, 2018, and included 879 patients with head and neck cancer who underwent pre-RT dental extractions before curative RT of 45 Gy or greater. Patient demographic information and clinical characteristics (eg, primary cancer site, nodal involvement, chemotherapy, smoking status, dental pathology) were considered. Data analyses were performed from July to December 2022. Timing (number of days) from dental extractions to RT start date and pre-RT extractions categorized as healed, minor bone spicules (MBS), or ORN. The study population consisted of 879 patients with a median (range) age of 62 (20-96) years, with 685 men (78%) and 194 women (22%). Of these, 847 (96.3%) healed from pre-RT dental extractions, 16 (1.8%) developed MBS, and 16 (1.8%) developed ORN. The median (range) time in number of days from pre-RT extraction(s) to start of RT was 9 (0-98) days in the healed cohort, 6 (3-23) days in the MBS cohort, and 6 (0-12) days in the ORN cohort. There was a large difference in the timing of pre-RT extractions between the healed and the MBS cohorts (mean 11.9 vs 7.4 days to radiation; difference 4.4; 95% CI, 1.5-7.3), and the healed and the ORN cohorts (mean 11.9 vs 7.1 days; difference 4.8 days; 95% CI, 2.6-7.1). The findings of this retrospective cohort study suggest that there was an important association between the timing of pre-RT dental extractions and ORN when extractions occurred within 7 days of the RT start date. Despite this, ORN after pre-RT extractions is relatively rare. These findings indicate that patients with head and neck cancer who are to undergo RT should not delay treatment for extractions when it might compromise oncologic control.

Quantitative and qualitative impact of physician assistants in a Canadian urology setting.

Physician assistants (PAs) are healthcare professionals who act as physician extenders. PAs are being used more and more in a wide variety of clinic settings throughout Canada to increase access to healthcare and reduce cost. We set out to determine the impact of PAs on a tertiary care center urologic oncology practice. We reviewed Ontario Health Insurance Plan (OHIP) billing codes since the introduction of PAs for two attending urologists at Princess Margaret Cancer Centre. Data were grouped into early experience and established experience. In addition, questionnaires were electronically distributed among nurses, physicians, residents, and fellows who work with PAs in clinic. Patient visits conducted by PAs were tracked for one quarter to estimate the amount of annual patients seen by PAs. The costs associated with PAs are presented as recommendations for a new graduate PA hire. On average, PAs increased clinic volume by 11.3 patient visits per day. Furthermore, they individually care for an average of 24 patients per day. PAs did not represent a financial burden on the urology practice plan (revenue gain of $16 800). Our questionnaire demonstrated that PAs were capable healthcare professionals, who decreased workload and contributed to resident/fellow education. PAs in a Canadian urology practice allow for more patient visits, decrease in physician workload, and positively impact trainee education. PAs saw more patients in clinic than clinic growth, thereby decreasing physician, fellow, and resident workload. The offset of the increase in patient visits made the PAs a cost-neutral investment.

Gestational trophoblastic neoplasia: does centralization of care impact clinical management?

ObjectiveInternational societies advocate for gestational trophoblastic neoplasia referral to designated expert centers. This study assessed the impact of centralization of trophoblastic care on clinical outcomes.MethodsA centralized program was implemented in...

Pediatric Radiation Therapy at the Princess Margaret Cancer Centre

Pediatric Radiation Therapy at the Princess Margaret Cancer Centre

Limitations of Prostate Biopsy in Detection of Cribriform and Intraductal Prostate Cancer

BackgroundThe presence of cribriform morphology and intraductal carcinoma (IDC) in prostate biopsies and radical prostatectomy specimens is an adverse prognostic feature that can be used to guide treatment decisions. ObjectiveTo assess how accurately biopsies can detect cribriform morphology and IDC cancer by examining matched biopsy and prostatectomy samples. Design, setting, and participantsPatients who underwent radical prostatectomy at The Princess Margaret Cancer Centre between January 2015 and December 2022 and had cribriform morphology and/or IDC in the surgical specimen were included in the study. Outcome measurements and statistical analysisWe used detection sensitivity to evaluate the level of agreement between biopsy and prostatectomy samples regarding the presence of cribriform morphology and IDC. Results and limitationsOf the 287 men who underwent radical prostatectomy, 241 (84%) had cribriform morphology and 161 (56%) had IDC on final pathology. The sensitivity of prostate biopsy, using radical prostatectomy as the reference, was 42.4% (95% confidence interval [CI] 36–49%) for detection of cribriform morphology and 44.1% (95% CI 36–52%) for detection of IDC. The sensitivity of prostate biopsy for detection of either IDC or cribriform morphology was 52.5% (95% CI 47–58%). Among patients who underwent multiparametric magnetic resonance imaging–guided biopsies, the sensitivity was 54% (95% CI 39–68%) for detection of cribriform morphology and 37% (95% CI 19–58%) for detection of IDC. ConclusionsBiopsy has low sensitivity for detecting cribriform morphology and IDC. These limitations should be incorporated into clinical decision-making. Biomarkers for better detection of these histological patterns are needed. Patient summaryProstate biopsy is not an accurate method for detecting two specific types of prostate cancer cells, called cribriform pattern and intraductal prostate cancer, which are associated with unfavorable prognosis.

Single centre retrospective analysis of extracorporeal photopheresis (ECP) therapy in patients heavily pre-treated for chronic graft-versus-host disease (cGvHD) with steroid failure

BackgroundExtracorporeal photopheresis (ECP) is recommended as a second- or later-line therapy for chronic GvHD (cGvHD). Benefits include reasonable response with avoidance of intense systemic immunosuppression, which can translate into lowering the risk of systemic toxicity and opportunistic infection. MethodsWe evaluated 75 patients treated with ECP for cGvHD from 2007 to 2021 at Princess Margaret Cancer Centre, and analyzed overall response rate (ORR) and clinical benefit (CB) at 3, 6 and 12 months plus other long-term treatment outcomes. ResultsWith a median follow-up of 72 months, a gradual increase in ORR was noted over time: 21% (16 out of 75 patients), 57% (36/63) and 70% (32/46) at month 3, 6 and 12, respectively. Gradual increase in CB was also observed over time with CB rate of 23% (17/75), 62% (39/63), and 76% (35/46) at months 3, 6 and 12, respectively. A total of 27 failures (36%) were noted, due to: 1) ECP resistance requiring switch to other therapy (n = 14, 19%), 2) non-relapse mortality (n = 10, 13%), 3) relapse of primary disease (n = 1, 1%) or 4) ECP procedure-related complication (n = 1, 1%, line infection), with 20 deaths (27%) observed. Failure-free survival (FFS) and overall survival (OS) rates were 68.3% and 85.9% at 12 months, respectively. After starting ECP, the proportions of patients who completely discontinued steroids were 17%, 32%, and 64% at months 3, 6 and 12, respectively. ConclusionECP is an effective treatment option for heavily pre-treated cGvHD patients.

Weight loss post-allogeneic stem cell transplant is associated with increased transplant-related mortality.

Allogeneic stem cell transplant (allo-HSCT) patients are at risk of malnutrition and weight loss from impaired oral intake resulting from gastrointestinal toxicities, dysgeusia, and psychological effects. A retrospective review of 264 adult patients transplanted at Princess Margaret Cancer Centre who achieved relapse-free survival up to 3months after allo-HSCT was performed. Overall incidence of patients who experienced WL (WL) ≥ 10% from HSCT to 3-month post-transplant was 45.9% and from HSCT to 6months was 56.6%. Patients with ≥ 10% WL from allo-HSCT at 3months and 6months had similar 2-year overall survival (OS) compared to those with < 10% WL, 55.7% vs 62.8% (HR = 1.38, p = 0.11) and 71.1% vs 77.2% (HR = 1.37, p = 0.27), respectively. Patients with ≥ 10% WL 3 and 6months from allo-HSCT also had similar 2-year relapse-free survival (RFS) compared to those with < 10% WL, 48.1% vs 55.8% (HR = 1.26, p = 0.22), and 62.7% vs 69.8% (HR = 1.29, p = 0.31), respectively. The 2-year transplant-related mortality (TRM) was higher for those with ≥ 10% WL from allo-HSCT to 3months, 35.4% vs 16.9% (HR = 2.39, p = 0.0007) and 6months, 22% vs 8% (HR = 3.1, p = 0.0034). Although statistical significance was not observed for OS or RFS, patients who experienced ≥ 10% WL 3- and 6-months post allo-HSCT experienced higher 2-year TRM. These results highlight the importance of early intervention and close monitoring of weight post allo-HSCT. Approaches to WL post allo-HSCT should be multifaceted and include members of the interdisciplinary team in order to decrease TRM.

A pan-cancer clinical platform to predict immunotherapy outcomes and prioritize immuno-oncology combinations in early-phase trials

Immunotherapy is effective, but current biomarkers forpatient selection have proven modest sensitivity. Here, we developed VIGex, an optimized gene signature based on the expression level of 12 genes involved in immune response with RNA sequencing. We implemented VIGex using the nCounter platform (Nanostring) on a large clinical cohort encompassing 909 tumor samples across 45 tumor types. VIGex was developed as a continuous variable, with cutoffs selected to detect three main categories (hot, intermediate-cold and cold) based on the different inflammatory status of the tumor microenvironment. Hot tumors had the highest VIGex scores and exhibited an increased abundance of tumor-infiltrating lymphocytes as compared with the intermediate-cold and cold. VIGex scores varied depending on tumor origin and anatomic site of metastases, with liver metastases showing an immunosuppressive tumor microenvironment. The predictive power of VIGex-Hot was observed in a cohort of 98 refractory solid tumor from patients treated in early-phase immunotherapy trials and its clinical performance was confirmed through an extensive metanalysis across 13 clinically annotated gene expression datasets from 877 patients treated with immunotherapy agents. Last, we generated a pan-cancer biomarker platform that integrates VIGex categories with the expression levels of immunotherapy targets under development in early-phase clinical trials. Our results support the clinical utility of VIGex as a tool toaid clinicians for patient selection and personalized immunotherapy interventions. BBVA Foundation; 202-2021 Division of Medical Oncology and Hematology Fellowship award; Princess Margaret Cancer Center.

Very late relapse in Hodgkin lymphoma: Characterizing an understudied population

BackgroundVery late relapse (VLR) occurring >5 years after initial diagnosis is an uncommon event in the management of Hodgkin lymphoma (HL). Limited information regarding risk factors and optimal therapy is available. Patients and MethodsWe reviewed patients treated for HL at Princess Margaret Cancer Centre, Toronto, Ontario Canada between January 01, 1999 and 31 December 31, 2018. ResultsThirty-two patients experienced VLR. Median time to first relapse was 7.2 years. Most patients were treated with CMT both at initial diagnosis and relapse. Male gender (P = .04) and increased age at initial diagnosis (P = .008; HR 1.09 (95% CI: 1.02-1.15)) were identified as risk factors for inferior survival on univariate analysis. Stage, histology, treatment modality and risk assessment at diagnosis or relapse did not have a significant impact on survival outcomes. ASCT at first relapse had no impact on time to second progression (HR 1.72; 95% CI, 0.35-8.53; P = .51) or overall survival from first relapse (HR 1.55; 95% CI, 0.3-8.03; P = .6). ConclusionOur data aligns with the limited information available in VLR HL suggesting the negative impact of age and male gender on this rare event. Additionally, our data did not show benefit of ASCT at first relapse in terms of survival outcomes in this population, though this analysis is limited by small sample size. Further study of optimal therapy to prevent and treat VL in the era of novel agents is critical.