SurvdigitizeR: an algorithm for automated survival curve digitization.

BACKGROUND: Decision analytic models and meta-analyses often rely on survival probabilities that are digitized from published Kaplan-Meier (KM) curves. However, manually extracting these probabilities from KM curves is time-consuming, expensive, and error-prone. We developed an efficient and accurate algorithm that automates extraction of survival probabilities from KM curves. METHODS: The automated digitization algorithm processes images from a JPG or PNG format, converts them in their hue, saturation, and lightness scale and uses optical character recognition to detect axis location and labels. It also uses a k-medoids clustering algorithm to separate multiple overlapping curves on the same figure. To validate performance, we generated survival plots form random time-to-event data from a sample size of 25, 50, 150, and 250, 1000 individuals split into 1,2, or 3 treatment arms. We assumed an exponential distribution and applied random censoring. We compared automated digitization and manual digitization performed by well-trained researchers. We calculated the root mean squared error (RMSE) at 100-time points for both methods. The algorithm's performance was also evaluated by Bland-Altman analysis for the agreement between automated and manual digitization on a real-world set of published KM curves. RESULTS: The automated digitizer accurately identified survival probabilities over time in the simulated KM curves. The average RMSE for automated digitization was 0.012, while manual digitization had an average RMSE of 0.014. Its performance was negatively correlated with the number of curves in a figure and the presence of censoring markers. In real-world scenarios, automated digitization and manual digitization showed very close agreement. CONCLUSIONS: The algorithm streamlines the digitization process and requires minimal user input. It effectively digitized KM curves in simulated and real-world scenarios, demonstrating accuracy comparable to conventional manual digitization. The algorithm has been developed as an open-source R package and as a Shiny application and is available on GitHub: https://github.com/Pechli-Lab/SurvdigitizeR and https://pechlilab.shinyapps.io/SurvdigitizeR/ .

Tightly integrated multiomics-based deep tensor survival model for time-to-event prediction.

MOTIVATION: Multiomics cancer profiles provide essential signals for predicting cancer survival. It is challenging to reveal the complex patterns from multiple types of data and link them to survival outcomes. We aim to develop a new deep learning-based algorithm to integrate three types of high-dimensional omics data measured on the same individuals to improve cancer survival outcome prediction. RESULTS: We built a three-dimension tensor to integrate multi-omics cancer data and factorized it into two-dimension matrices of latent factors, which were fed into neural networks-based survival networks. The new algorithm and other multi-omics-based algorithms, as well as individual genomic-based survival analysis algorithms, were applied to the breast cancer data colon and rectal cancer data from The Cancer Genome Atlas (TCGA) program. We evaluated the goodness-of-fit using the concordance index (C-index) and Integrated Brier Score (IBS). We demonstrated that the proposed tight integration framework has better survival prediction performance than the models using individual genomic data and other conventional data integration methods. AVAILABILITY AND IMPLEMENTATION: https://github.com/jasperzyzhang/DeepTensorSurvival. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Clinical outcomes of oral epithelial dysplasia managed by observation versus excision.

IMPORTANCE: Oral potentially malignant disorders, including oral epithelial dysplasia (OED), are a group of conditions with an increased risk of progression to oral cancer. Clinical management of OED is challenging and usually involves monitoring with repeated incisional biopsies or complete surgical excision. OBJECTIVE: To determine if complete surgical excision of OED impacts malignant transformation or improves survival outcomes in lesions that progress to malignancy. DESIGN: A retrospective review of all patients diagnosed with OED between 2009 and 2016 was completed, and patients were followed until January 2022 for disease course and outcomes. RESULTS: Hundred and fifty-five cases of OED met the inclusion criteria. Among the 61 lesions managed by observation, 15 progressed to cancer. Among the 94 lesions managed by surgical excision, 27 progressed to cancer. The overall malignant transformation rate was 27%, with an annual rate of 6.4%. Surgical excision with or without histologically negative margins did not decrease malignant transformation but was associated with lower oncologic staging at the time of diagnosis and improved survival. CONCLUSIONS AND RELEVANCE: Surgical excision of OED with or without negative margins did not reduce the rate of transformation to oral cancer but resulted in lower oncologic staging at diagnosis, leading to improved patient outcomes. Our results support the implementation of more extensive tissue sampling to improve cancer diagnosis and patient outcomes.

Very late relapse in Hodgkin lymphoma: Characterizing an understudied population.

BACKGROUND: Very 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 METHODS: We reviewed patients treated for HL at Princess Margaret Cancer Centre, Toronto, Ontario Canada between January 01, 1999 and 31 December 31, 2018. RESULTS: Thirty-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). CONCLUSION: Our 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.