Application of Machine Learning to the Prediction of Cancer-Associated Venous Thromboembolism.

Venous thromboembolism (VTE) is a common and impactful complication of cancer. Several clinical prediction rules have been devised to estimate the risk of a thrombotic event in this patient population, however they are associated with limitations. We aimed to develop a predictive model of cancer-associated VTE using machine learning as a means to better integrate all available data, improve prediction accuracy and allow applicability regardless of timing for systemic therapy administration. A retrospective cohort was used to fit and validate the models, consisting of adult patients who had next generation sequencing performed on their solid tumor for the years 2014 to 2019. A deep learning survival model limited to demographic, cancer-specific, laboratory and pharmacological predictors was selected based on results from training data for 23,800 individuals and was evaluated on an internal validation set including 5,951 individuals, yielding a time-dependent concordance index of 0.72 (95% CI = 0.70-0.74) for the first 6 months of observation. Adapted models also performed well overall compared to the Khorana Score (KS) in two external cohorts of individuals starting systemic therapy; in an external validation set of 1,250 patients, the C-index was 0.71 (95% CI = 0.65-0.77) for the deep learning model vs 0.66 (95% CI = 0.59-0.72) for the KS and in a smaller external cohort of 358 patients the C-index was 0.59 (95% CI = 0.50-0.69) for the deep learning model vs 0.56 (95% CI = 0.48-0.64) for the KS. The proportions of patients accurately reclassified by the deep learning model were 25% and 26% respectively. In this large cohort of patients with a broad range of solid malignancies and at different phases of systemic therapy, the use of deep learning resulted in improved accuracy for VTE incidence predictions. Additional studies are needed to further assess the validity of this model.

Relations of Current and Past Cancer with Severe Outcomes among 104,590 Hospitalized COVID-19 Patients: The COVID EHR Cohort at the University of Wisconsin.

BACKGROUND: There is mixed evidence about the relations of current versus past cancer with severe COVID-19 outcomes and how they vary by patient and cancer characteristics. METHODS: Electronic health record data of 104,590 adult hospitalized patients with COVID-19 were obtained from 21 United States health systems from February 2020 through September 2021. In-hospital mortality and ICU admission were predicted from current and past cancer diagnoses. Moderation by patient characteristics, vaccination status, cancer type, and year of the pandemic was examined. RESULTS: 6.8% of the patients had current (n = 7,141) and 6.5% had past (n = 6,749) cancer diagnoses. Current cancer predicted both severe outcomes but past cancer did not; adjusted odds ratios (aOR) for mortality were 1.58 [95% confidence interval (CI), 1.46-1.70] and 1.04 (95% CI, 0.96-1.13), respectively. Mortality rates decreased over the pandemic but the incremental risk of current cancer persisted, with the increment being larger among younger vs. older patients. Prior COVID-19 vaccination reduced mortality generally and among those with current cancer (aOR, 0.69; 95% CI, 0.53-0.90). CONCLUSIONS: Current cancer, especially among younger patients, posed a substantially increased risk for death and ICU admission among patients with COVID-19; prior COVID-19 vaccination mitigated the risk associated with current cancer. Past history of cancer was not associated with higher risks for severe COVID-19 outcomes for most cancer types. IMPACT: This study clarifies the characteristics that modify the risk associated with cancer on severe COVID-19 outcomes across the first 20 months of the COVID-19 pandemic. See related commentary by Egan et al., p. 3.

Effectiveness of MRNA booster vaccine among healthcare workers in New York City during the Omicron surge, December 2021 to January 2022.

OBJECTIVE: To describe effectiveness of mRNA vaccines by comparing 2-dose (2D) and 3-dose (3D) healthcare worker (HCW) recipients in the setting of Omicron variant dominance. Performance of 2D and 3D vaccine series against SARS-CoV-2 variants and the clinical outcomes of HCWs may inform return-to-work guidance. METHODS: In a retrospective study from December 15, 2020 to January 15, 2022, SARS-CoV-2 infections among HCWs at a large tertiary cancer centre in New York City were examined to estimate infection rates (aggregated positive tests / person-days) and 95% CIs over the Omicron period in 3D and 2D mRNA vaccinated HCWs and were compared using rate ratios. We described the clinical features of post-vaccine infections and impact of prior (pre-Omicron) COVID infection on vaccine effectiveness. RESULTS: Among the 20857 HCWs in our cohort, 20,660 completed the 2D series with an mRNA vaccine during our study period and 12461 had received a third dose by January 15, 2022. The infection rate ratio for 3D versus 2D vaccinated HCWs was 0.667 (95% CI 0.623, 0.713) for an estimated 3D vaccine effectiveness of 33.3% compared to two doses only during the Omicron dominant period from December 15, 2021 to January 15, 2022. Breakthrough Omicron infections after 3D + 14 days occurred in 1,315 HCWs. Omicron infections were mild, with 16% of 3D and 11% 2D HCWs being asymptomatic. DISCUSSION: Study demonstrates improved vaccine-derived protection against COVID-19 infection in 3D versus 2D mRNA vaccinees during the Omicron surge. The advantage of 3D vaccination was maintained irrespective of prior COVID-19 infection status.

Implementing a Clinical Decision Tool to Improve Oncologic Venous Thromboembolism Management.

Background: Cancer patients with venous thromboembolic (VTE) disease are complex, and many factors must be considered when initiating anticoagulation management. Clinical decision support systems can aid in decision-making by utilizing guidelines at the point of care. Objectives: The purpose of our project was to develop, implement, and evaluate an electronic clinical decision tool (CDT) utilizing evidence-based guidelines to aid in decision-making for adult oncologic patients who present with new VTE to symptom care clinics. Methods: We compared a pre-intervention group of patients who were prescribed anticoagulation (n = 98) with two post-intervention groups: CDT applied (n = 96) and not applied (n = 46). Outcomes included whether the CDT anticoagulation recommendations were followed and if the tool was perceived to be helpful or improve confidence in initiating management for new VTE by the SCC advanced practitioners and physicians. Results: There was no significant difference between the pre- and post-intervention groups in how many of the CDT anticoagulation recommendations were followed (68.8% pre-intervention, 60.8% post-intervention tool applied, and 63.5% post-intervention tool not applied; χ2 [2, N = 161] = .921, p = .631). However, the tool was found to be helpful and improved confidence of the providers in initiating management for new VTE (pre median = 3, interquartile range [IQR] = 2, 3.5; post median = 3, IQR 3, 4; p = .033). Conclusion: This CDT provided evidence-based anticoagulation recommendations for cancer-associated VTE and enhanced familiarity with the standard of care. Further development of the CDT will be required to account for situations that resulted in deviation from the recommendations.

Clinical and Genomic Characterization of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS CoV-2) Infections in mRNA Vaccinated Health Care Personnel in New York City.

BACKGROUND: Vaccine-induced clinical protection against severe acute respiratory syndrome coronavirus 2 (SARS CoV-2) variants is an evolving target. There are limited genomic level data on SARS CoV-2 breakthrough infections and vaccine effectiveness (VE) since the global spread of the B.1.617.2 (Delta) variant. METHODS: In a retrospective study from 1 November 2020 to 31 August 2021, divided as pre-Delta and Delta-dominant periods, laboratory-confirmed SARS CoV-2 infections among healthcare personnel (HCP) at a large tertiary cancer center in New York City were examined to compare the weekly infection rate-ratio in vaccinated, partially vaccinated, and unvaccinated HCP. We describe the clinical and genomic epidemiologic features of post-vaccine infections to assess for selection of variants of concern (VOC)/variants of interest (VOI) in the early post-vaccine period and impact of B.1.617.2 (Delta) variant domination on VE. RESULTS: Among 13658 HCP in our cohort, 12379 received at least 1 dose of a messenger RNA (mRNA) vaccine. In the pre-Delta period overall VE was 94.5%. Whole genome sequencing (WGS) of 369 isolates in the pre-Delta period did not reveal a clade bias for VOC/VOI specific to post-vaccine infections. VE in the Delta dominant phase was 75.6%. No hospitalizations occurred among vaccinated HCP in the entire study period, compared to 17 hospitalizations and 1 death among unvaccinated HCP. CONCLUSIONS: Findings show high VE among HCP in New York City in the pre-Delta phase, with moderate decline in VE post-Delta emergence. SARS CoV-2 clades were similarly distributed among vaccinated and unvaccinated infected HCP without apparent clustering during the pre-Delta period of diverse clade circulation. Strong vaccine protection against hospitalization was maintained through the entire study period.