Clinical variables associated with immune checkpoint inhibitor outcomes in patients with metastatic urothelial carcinoma: a multicentre retrospective cohort study.

OBJECTIVES: Immune checkpoint inhibitors (ICIs) are indicated for metastatic urothelial cancer (mUC), but predictive and prognostic factors are lacking. We investigated clinical variables associated with ICI outcomes. METHODS: We performed a multicentre retrospective cohort study of 135 patients who received ICI for mUC, 2016-2021, at three Canadian centres. Clinical characteristics, body mass index (BMI), metastatic sites, neutrophil-to-lymphocyte ratio (NLR), response and survival were abstracted from chart review. RESULTS: We identified 135 patients and 62% had received ICI as a second-line or later treatment for mUC. A BMI ≥25 was significantly correlated to a higher overall response rate (ORR) (45.4% vs 16.3%, p value=0.020). Patients with BMI ≥30 experienced longer median overall survival (OS) of 24.8 vs 14.4 for 25≤BMI<30 and 8.5 months for BMI <25 (p value=0.012). The ORR was lower in the presence of bone metastases (16% vs 41%, p value=0.006) and liver metastases (16% vs 39%, p value=0.013). Metastatic lymph nodes were correlated with higher ORR (40% vs 20%, p value=0.032). The median OS for bone metastases was 7.3 versus 18 months (p value <0.001). Patients with liver metastases had a median OS of 8.6 versus 15 months (p value=0.006). No difference for lymph nodes metastases (13.5 vs 12.7 months, p value=0.175) was found. NLR ≥4 had worse OS (8.2 vs 17.7 months, p value=0.0001). In multivariate analysis, BMI ≥30, bone metastases, NLR ≥4, performance status ≥2 and line of ICI ≥2 were independent factors for OS. CONCLUSIONS: Our data identified BMI and bone metastases as novel clinical biomarkers that were independently associated with ICI outcomes in mUC. External and prospective validation are warranted.

Recommended first-line management of asymptomatic brain metastases from EGFR mutant and ALK positive non-small cell lung cancer varies significantly according to specialty: an international survey of clinical practice.

Background: The role for radiotherapy or surgery in the upfront management of brain metastases (BrM) in epidermal growth factor receptor mutant (EGFRm) or anaplastic lymphoma kinase translocation positive (ALK+) non-small cell lung cancer (NSCLC) is uncertain because of a lack of prospective evidence supporting tyrosine kinase inhibitor (TKI) monotherapy. Further understanding of practice heterogeneity is necessary to guide collaborative efforts in establishing guideline recommendations. Methods: We conducted an international survey among medical (MO), clinical (CO), and radiation oncologists (RO), as well as neurosurgeons (NS), of treatment recommendations for asymptomatic BrM (in non-eloquent regions) EGFRm or ALK+ NSCLC patients according to specific clinical scenarios. We grouped and compared treatment recommendations according to specialty. Responses were summarized using counts and percentages and analyzed using the Fisher exact test. Results: A total of 449 surveys were included in the final analysis: 48 CO, 85 MO, 60 NS, and 256 RO. MO and CO were significantly more likely than RO and NS to recommend first-line TKI monotherapy, regardless of the number and/or size of asymptomatic BrM (in non-eloquent regions). Radiotherapy in addition to TKI as first-line management was preferred by all specialties for patients with ≥4 BrM. NS recommended surgical resection more often than other specialties for BrM measuring >2 cm. Conclusions: Recommendations for the management of BrM from EGFRm or ALK+ NSCLC vary significantly according to oncology sub-specialties. Development of multidisciplinary guidelines and further research on establishing optimal treatment strategies is warranted.

Three-year disease-free survival in randomized trials of neoadjuvant chemotherapy and HER2-targeted therapy in breast cancer: A meta-analysis.

BACKGROUND: Outcomes for breast cancer patients with residual disease (RD) after neoadjuvant chemotherapy (NACT) and HER2-targeted therapy may be better than anticipated leading to a smaller absolute benefit of adjuvant trastuzumab emtansine (T-DM1). Therefore, accurate estimates of 3-year disease-free survival (DFS) can aid in treatment planning. METHODS: We reviewed randomized trials of NACT and HER2-targeted therapy in breast cancer (excluding T-DM1) and calculated mean 3-year DFS weighted by study sample size. Meta-regression comprising linear regression weighted by sample size (mixed-effects) was performed to explore associations between 3-year DFS and year of accrual and trial-level patient, disease, and treatment factors. Data were reported quantitatively irrespective of statistical significance. RESULTS: Eleven studies (N = 3581) were included in the primary analysis. The mean 3-year DFS for patients with RD was 79.7% (95% CI 77.4-80.9). This was higher for trials completing accrual after 2010 [83% (95% CI 79.3-86.3)] and for those receiving dual HER2 targeted therapy [83.4% (95% CI 79.2-87.7]. Better outcomes for ER positivity, later accrual and dual Her-2 targeted therapy were confirmed in meta-regression. Negative quantitative significance was observed for larger clinical tumor size and nodal involvement. CONCLUSIONS: The 3-year DFS for patients with RD has improved over time possibly due to dual HER2 targeted therapy. This will reduce the absolute benefit of adjuvant T-DM1 in this group of patients.

Cell-free DNA methylation-defined prognostic subgroups in small-cell lung cancer identified by leukocyte methylation subtraction.

Small-cell lung cancer (SCLC) methylome is understudied. Here, we comprehensively profile SCLC using cell-free methylated DNA immunoprecipitation followed by sequencing (cfMeDIP-seq). Cell-free DNA (cfDNA) from plasma of 74 patients with SCLC pre-treatment and from 20 non-cancer participants, genomic DNA (gDNA) from peripheral blood leukocytes from the same 74 patients, and 7 accompanying circulating tumor cell-derived xenografts (CDXs) underwent cfMeDIP-seq. Peripheral blood leukocyte methylation (PRIME) subtraction to improve tumor specificity. SCLC cfDNA methylation is distinct from non-cancer but correlates with CDX tumor methylation. PRIME and k-means consensus identified two methylome clusters with prognostic associations that related to axon guidance, neuroactive ligand-receptor interaction, pluripotency of stem cells, and differentially methylated at long noncoding RNA and other repeats features. We comprehensively profiled the SCLC methylome in a large patient cohort and identified methylome clusters with prognostic associations. Our work demonstrates the potential of liquid biopsies in examining SCLC biology encoded in the methylome.

Prognostic Value of Programmed Death Ligand-1 Expression in Solid Tumors Irrespective of Immunotherapy Exposure: A Systematic Review and Meta-Analysis.

BACKGROUND: The programmed cell death-1/programmed cell death ligand-1 (PD-L1) pathway, which plays a crucial role in cancer immune surveillance, is the target of several approved immunotherapeutic agents and is used as a predictive biomarker in some solid tumors. However, its use as a prognostic marker (i.e., regardless of therapy used) is not established clearly with available data demonstrating inconsistent prognostic impact of PD-L1 expression in solid tumors. METHODS: We conducted a systematic literature search of electronic databases and identified publications exploring the effect of PD-L1 expression on overall survival and/or disease-free survival. Hazard ratios were pooled in a meta-analysis using generic inverse-variance and random-effects modeling. We used the Deeks method to explore subgroup differences based on disease site, stage of disease, and method of PD-L1 quantification. RESULTS: One hundred and eighty-six studies met the inclusion criteria. Programmed cell death ligand-1 expression was associated with worse overall survival (hazard ratio 1.33, 95% confidence interval 1.26-1.39; p < 0.001). There was significant heterogeneity between disease sites (subgroup p = 0.002) with pancreatic, hepatocellular, and genitourinary cancers associated with the highest magnitude of adverse outcomes. Programmed cell death ligand-1 was also associated with worse overall disease-free survival (hazard ratio 1.19, 95% confidence interval 1.09-1.30; p < 0.001). Stage of disease did not significantly affect the results (subgroup p = 0.52), nor did the method of quantification via immunohistochemistry or messenger RNA (subgroup p = 0.70). CONCLUSIONS: High expression of PD-L1 is associated with worse survival in solid tumors albeit with significant heterogeneity among tumor types. The effect is consistent in early-stage and metastatic disease and is not sensitive to method of PD-L1 quantification. These data can provide additional information for the counseling of patients with cancer about prognosis.

Delivery of Virtual Care in Oncology: Province-Wide Interprofessional Consensus Statements Using a Modified Delphi Process.

Virtual cancer care (i.e., teleoncology) was rapidly adopted during the COVID-19 pandemic to meet the needs of patients with cancer. However, there is a paucity of guidance for clinicians regarding virtual cancer care. We sought to develop consensus-based statements to guide the optimal provision of virtual care for clinicians caring for patients with cancer, using a modified Delphi consensus process with a 29-member panel consisting of an interprofessional group of clinicians caring for patients with cancer and patient representatives. The consensus process consisted of two rounds and one synchronous final consensus meeting. At the end of the modified Delphi process, 62 of 62 statements achieved consensus. Fifty-seven statements reached consensus in the first round of the process. Concerns regarding the ability to convey difficult news virtually and maintaining similar standards as in-person care without disproportionate strain on clinicians and patients were addressed in the consensus process. We achieved interprofessional consensus on virtual cancer care practices. Further research examining the impact of virtual cancer care on person-centred and clinical outcomes are needed to inform practices during the COVID-19 pandemic and beyond.

Genomic sequencing to inform therapy in advanced pancreatic cancer: A systematic review and meta-analysis of prospective studies.

PURPOSE: Current guidelines recommend somatic genomic sequencing for patients with advanced pancreatic cancer to identify targetable alterations amenable to targeted therapy. The benefit of somatic genomic sequencing in pancreatic cancer remains unclear. This study aims to assess the evidence supporting genomic sequencing to inform treatment selection for patients with advanced pancreatic cancer. METHODS: A systematic review identified prospective studies of exocrine pancreatic cancer patients published before August 2020 which conducted genomic sequencing to inform treatment selection. Outcomes of interest included the proportion of patients with targetable alterations, the proportion that received targeted treatments, and the impact of targeted treatments on overall survival. Meta-analysis for proportions and hazard ratios was performed using Dersimonian and Laird random effect models. RESULTS: 19 studies (representing 2048 pancreatic cancer patients) were included. Sequencing methodologies, definitions of targetable alterations, and approaches treatment selection varied across studies and were incompletely reported. 590 of 1382 sequenced patients harboured a targetable alteration (random effects meta-analysis estimate of the proportion 0.46, 95% confidence interval 0.32-0.61). The proportion of patients with targetable alterations was highly heterogenous between studies (I2 93%, P < 0.001). 91 of 1390 patients received a matched therapy based on their targetable alterations (random effects meta-analysis estimate of the proportion 0.12, 95% CI 0.06-0.23). One observational study reported an overall survival benefit of matched therapy. CONCLUSIONS: Genomic sequencing frequently identifies targetable alterations in pancreatic cancers. Further research is required to standardize the definitions of targetable alterations, the approach to treatment matching, and quantify the benefit of targeted therapy.

Cancer Treatment During COVID-19: A Qualitative Analysis of Patient-Perceived Risks and Experiences with Virtual Care.

During the COVID-19 pandemic, most cancer centers shifted from in-person to virtual cancer care to curb community spread and ensure care continuity. This qualitative descriptive study aimed to understand cancer patient-perceived risks related to COVID-19 and cancer treatment, as well as the patient-perceived and experienced value of virtual care. From June to August 2020, focus groups were conducted with patients under active management or observation for a diagnosed malignancy in Toronto, Canada. A thematic analysis of six focus groups found that most participants worried more about treatment delays than they did about COVID-19 infection. Despite some concern about COVID-19 exposure in the hospital, care delays contributed to increased anxiety among participants who already subscribed to strict safety measures in their everyday lives. Most participants accepted virtual care for some appointment types; however, preference for in-person care was found to sustain the humanistic and therapeutic aspects of cancer care that many participants valued. Nuances in the appropriateness and adequacy of virtual cancer care still need exploration. Preserving the humanistic aspects of care is of paramount importance.

Platinum-based chemotherapy in early-stage triple negative breast cancer: A meta-analysis.

PURPOSE: The addition of platinum agents to anthracycline and taxane-based chemotherapy in early-stage triple negative breast cancer (TNBC) patients improves pathological complete response (pCR). Long-term outcomes, such as disease-free survival (DFS) and overall survival (OS), have not been well-established. METHODS: A systematic literature review identified studies using platinum-based treatment in TNBC patients in the neoadjuvant or adjuvant setting with reportable long-term outcomes. Hazard ratios (HR) from collected data were pooled in a meta-analysis using generic inverse-variance and random effects modeling. Subgroup analyses were conducted based on treatment setting and study design. RESULTS: Fourteen studies comprising 3518 patients met the inclusion criteria. Median follow up was 56.2 months. All studies reported DFS and 9 studies (64%) reported OS. DFS was significantly better in platinum-based treatment (HR 0.71, 95% confidence interval (CI) 0.56-0.89; p = 0.03). However, OS was no different (HR 0.98, 95% CI 0.75-1.27; p = 0.87). There was a non-significant difference between platinum exposure in the adjuvant compared to neoadjuvant setting for both DFS (HR 0.75 vs 0.62, p = 0.43) and for OS (HR 0.90 vs 1.10, p = 0.58). The addition of platinum was associated with more thrombocytopenia and all-grade neuropathy and non-significant increases in neutropenia and grade 3-4 neuropathy. CONCLUSIONS: Platinum-based treatment improves DFS but not OS. The reporting of toxicity was suboptimal, but in general adding platinum increased toxicity. The discordant effect of platinum-based treatment on DFS and OS suggest the potential development of platinum resistance and worse outcomes after recurrence. Platinum-based chemotherapy cannot be recommended in unselected patients with early TNBC.

Quality of patient-reported outcomes in oncology clinical trials using immune checkpoint inhibitors: A systematic review.

BACKGROUND: There are limited data regarding the quality of patient-reported outcome (PRO) data in immune checkpoint inhibitor (ICI) clinical trial publications. METHODS: A systematic search of citations from various databases was conducted to identify prospective clinical trials involving ICI in advanced tumors from 2003 to 2020. A 30-point score was adapted from the CONSORT PRO extension statement to assess adherence to CONSORT PRO reporting. Linear regression was used to identify factors associated with quality reporting. RESULTS: After the review of 8058 articles, 33 trials were included with ICIs as either monotherapy (91%) or part of a combination regimen (9%). The median score was 23.5 points (range 15-29). In the majority of cases (82%), PROs were reported in a separate publication from the original study. Most of the trials were conducted in the metastatic setting and predominantly in melanoma, lung, and renal cell carcinoma (RCC) (73%). Univariate analysis revealed that trials with greater than 250 patients were associated with a higher score. The score was more likely to be lower in disease sites other than melanoma, lung, and RCC and was higher in the KEYNOTE than in the CHECKMATE trial series. There was no significant correlation between the score and whether a trial met its primary end-point or if the trial improved or worsened the quality of life. In the multivariate analysis, the number of patients enrolled to the trial, disease site, and trial series remained significant. CONCLUSIONS: The quality of reporting of PROs in ICI phase II and III clinical trials is heterogeneous across various cancer sites. As PRO data are increasingly used to counsel patients and complement clinical decision making, innovative and collaborative efforts are required to improve the reporting of these essential data.

Claudin-2 promotes colorectal cancer liver metastasis and is a biomarker of the replacement type growth pattern.

Claudin-2 promotes breast cancer liver metastasis by enabling seeding and early cancer cell survival. We now demonstrate that Claudin-2 is functionally required for colorectal cancer liver metastasis and that Claudin-2 expression in primary colorectal cancers is associated with poor overall and liver metastasis-free survival. We have examined the role of Claudin-2, and other claudin family members, as potential prognostic biomarkers of the desmoplastic and replacement histopathological growth pattern associated with colorectal cancer liver metastases. Immunohistochemical analysis revealed higher Claudin-2 levels in replacement type metastases when compared to those with desmoplastic features. In contrast, Claudin-8 was highly expressed in desmoplastic colorectal cancer liver metastases. Similar observations were made following immunohistochemical staining of patient-derived xenografts (PDXs) that we have established, which faithfully retain the histopathology of desmoplastic or replacement type colorectal cancer liver metastases. We provide evidence that Claudin-2 status in patient-derived extracellular vesicles may serve as a relevant prognostic biomarker to predict whether colorectal cancer patients have developed replacement type liver metastases. Such a biomarker will be a valuable tool in designing optimal treatment strategies to better manage patients with colorectal cancer liver metastases.

A review and comparison of breast tumor cell nuclei segmentation performances using deep convolutional neural networks.

Breast cancer is currently the second most common cause of cancer-related death in women. Presently, the clinical benchmark in cancer diagnosis is tissue biopsy examination. However, the manual process of histopathological analysis is laborious, time-consuming, and limited by the quality of the specimen and the experience of the pathologist. This study's objective was to determine if deep convolutional neural networks can be trained, with transfer learning, on a set of histopathological images independent of breast tissue to segment tumor nuclei of the breast. Various deep convolutional neural networks were evaluated for the study, including U-Net, Mask R-CNN, and a novel network (GB U-Net). The networks were trained on a set of Hematoxylin and Eosin (H&E)-stained images of eight diverse types of tissues. GB U-Net demonstrated superior performance in segmenting sites of invasive diseases (AJI = 0.53, mAP = 0.39 & AJI = 0.54, mAP = 0.38), validated on two hold-out datasets exclusively containing breast tissue images of approximately 7,582 annotated cells. The results of the networks, trained on images independent of breast tissue, demonstrated that tumor nuclei of the breast could be accurately segmented.

Machine Learning Frameworks to Predict Neoadjuvant Chemotherapy Response in Breast Cancer Using Clinical and Pathological Features.

PURPOSE: Neoadjuvant chemotherapy (NAC) is used to treat locally advanced breast cancer (LABC) and high-risk early breast cancer (BC). Pathological complete response (pCR) has prognostic value depending on BC subtype. Rates of pCR, however, can be variable. Predictive modeling is desirable to help identify patients early who may have suboptimal NAC response. Here, we test and compare the predictive performances of machine learning (ML) prediction models to a standard statistical model, using clinical and pathological data. METHODS: Clinical and pathological variables were collected in 431 patients, including tumor size, patient demographics, histological characteristics, molecular status, and staging information. A standard multivariable logistic regression (MLR) was developed and compared with five ML models: k-nearest neighbor classifier, random forest (RF) classifier, naive Bayes algorithm, support vector machine, and multilayer perceptron model. Model performances were measured using a receiver operating characteristic (ROC) analysis and statistically compared. RESULTS: MLR predictors of NAC response included: estrogen receptor (ER) status, human epidermal growth factor-2 (HER2) status, tumor size, and Nottingham grade. The strongest MLR predictors of pCR included HER2+ versus HER2- BC (odds ratio [OR], 0.13; 95% CI, 0.07 to 0.23; P < .001) and Nottingham grade G3 versus G1-2 (G1-2: OR, 0.36; 95% CI, 0.20 to 0.65; P < .001). The area under the curve (AUC) for the MLR was AUC = 0.64. Among the various ML models, an RF classifier performed best, with an AUC = 0.88, sensitivity of 70.7%, and specificity of 84.6%, and included the following variables: menopausal status, ER status, HER2 status, Nottingham grade, tumor size, nodal status, and presence of inflammatory BC. CONCLUSION: Modeling performances varied between standard versus ML classification methods. RF ML classifiers demonstrated the best predictive performance among all models.

SARS-CoV-2 Testing for Asymptomatic Patients with Cancer Prior and during Treatment: A Single Centre Experience.

Patients with cancer are more vulnerable to severe COVID-19. As a result, routine SARS-CoV-2 testing of asymptomatic patients with cancer is recommended prior to treatment. However, there is limited evidence of its clinical usefulness. The objective of this study is to evaluate the value of routine testing of asymptomatic patients with cancer. Asymptomatic patients with cancer attending Odette Cancer Centre (Toronto, ON, Canada) were tested for SARS-CoV-2 prior to and during treatment cycles. Results were compared to positivity rates of SARS-CoV-2 locally and provincially. All 890 asymptomatic patients tested negative. Positivity rates in the province were 1.5%, in hospital were 1.0%, and among OCC's symptomatic cancer patients were 0% over the study period. Given our findings and the low SARS-CoV-2 community positivity rates, we recommend a dynamic testing model of asymptomatic patients that triggers testing during increasing community positivity rates of SARS-CoV-2.

Computational Radiology in Breast Cancer Screening and Diagnosis Using Artificial Intelligence.

Breast cancer screening has been shown to significantly reduce mortality in women. The increased utilization of screening examinations has led to growing demands for rapid and accurate diagnostic reporting. In modern breast imaging centers, full-field digital mammography (FFDM) has replaced traditional analog mammography, and this has opened new opportunities for developing computational frameworks to automate detection and diagnosis. Artificial intelligence (AI), and its subdomain of deep learning, is showing promising results and improvements on diagnostic accuracy, compared to previous computer-based methods, known as computer-aided detection and diagnosis.In this commentary, we review the current status of computational radiology, with a focus on deep neural networks used in breast cancer screening and diagnosis. Recent studies are developing a new generation of computer-aided detection and diagnosis systems, as well as leveraging AI-driven tools to efficiently interpret digital mammograms, and breast tomosynthesis imaging. The use of AI in computational radiology necessitates transparency and rigorous testing. However, the overall impact of AI to radiology workflows will potentially yield more efficient and standardized processes as well as improve the level of care to patients with high diagnostic accuracy.

Associations between safety, tolerability, and toxicity and the reporting of health-related quality of life in phase III randomized trials in common solid tumors.

BACKGROUND: Anti-cancer drugs are approved typically on the basis of efficacy and safety as evaluated in phase III randomized trials (RCTs). Health-related quality of life (HRQoL) is a direct measure of patient benefit, but is under-reported. Here we explore associations with reporting of HRQoL data in phase III RCTs in common solid tumors. METHODS: We searched ClinicalTrials.gov to identify phase III RCTs evaluating new drugs in adults with advanced cancers that completed accrual between January 2005 and October 2016. Data on HRQoL, safety, and tolerability comprising treatment-related death, treatment discontinuation and commonly reported grade 3 or 4 adverse events (AEs) were extracted. Associations between these measures and reporting of HRQoL data were explored using logistic regression. RESULTS: Of 377 phase III RCTs identified initially, 143 studies were analysed and comprised 55% positive trials and 90% industry sponsored trials. HRQoL was listed as an endpoint in 59% trials; and of these, only 65% reported HRQoL data. There were higher odds of reporting HRQoL data for positive trials (OR 2.05, P = .04) and trials published in journals with higher impact factor (OR 1.35, P = .01). Reporting of HRQoL was not associated with treatment-related death (OR 1.25, P = .40) or treatment discontinuation (OR 1.12, P = .61), but was positively associated with dyspnea and dermatological adverse events. CONCLUSIONS: HRQoL is reported in only two-thirds of RCTs that describe collecting such data. Reporting of HRQoL is associated with positive trial outcome and higher journal impact factor, but not associated with overall safety and tolerability of anti-cancer drugs.