Machine learning algorithms to estimate 10-Year survival in patients with bone metastases due to prostate cancer: toward a disease-specific survival estimation tool.

BACKGROUND: Prognostic indicators, treatments, and survival estimates vary by cancer type. Therefore, disease-specific models are needed to estimate patient survival. Our primary aim was to develop models to estimate survival duration after treatment for skeletal-related events (SREs) (symptomatic bone metastasis, including impending or actual pathologic fractures) in men with metastatic bone disease due to prostate cancer. Such disease-specific models could be added to the PATHFx clinical-decision support tool, which is available worldwide, free of charge. Our secondary aim was to determine disease-specific factors that should be included in an international cancer registry. METHODS: We analyzed records of 438 men with metastatic prostate cancer who sustained SREs that required treatment with radiotherapy or surgery from 1989-2017. We developed and validated 6 models for 1-, 2-, 3-, 4-, 5-, and 10-year survival after treatment. Model performance was evaluated using calibration analysis, Brier scores, area under the receiver operator characteristic curve (AUC), and decision curve analysis to determine the models' clinical utility. We characterized the magnitude and direction of model features. RESULTS: The models exhibited acceptable calibration, accuracy (Brier scores < 0.20), and classification ability (AUCs > 0.73). Decision curve analysis determined that all 6 models were suitable for clinical use. The order of feature importance was distinct for each model. In all models, 3 factors were positively associated with survival duration: younger age at metastasis diagnosis, proximal prostate-specific antigen (PSA) < 10 ng/mL, and slow-rising alkaline phosphatase velocity (APV). CONCLUSIONS: We developed models that estimate survival duration in patients with metastatic bone disease due to prostate cancer. These models require external validation but should meanwhile be included in the PATHFx tool. PSA and APV data should be recorded in an international cancer registry.

A Machine-Learning Algorithm to Predict the Likelihood of Prolonged Opioid Use Following Arthroscopic Hip Surgery.

PURPOSE: To develop a machine-learning algorithm and clinician-friendly tool predicting the likelihood of prolonged opioid use (>90 days) following hip arthroscopy. METHODS: The Military Data Repository was queried for all adult patients undergoing arthroscopic hip surgery between 2012 and 2017. Demographic, health history, and prescription records were extracted for all included patients. Opioid use was divided into preoperative use (30-365 days before surgery), perioperative use (30 days before surgery through 14 days after surgery), postoperative use (14-90 days after surgery), and prolonged postoperative use (90-365 days after surgery). Six machine-learning algorithms (Naïve Bayes, Gradient Boosting Machine, Extreme Gradient Boosting, Random Forest, Elastic Net Regularization, and artificial neural network) were developed. Area under the receiver operating curve and Brier scores were calculated for each model. Decision curve analysis was applied to assess clinical utility. Local-Interpretable Model-Agnostic Explanations were used to demonstrate factor weights within the selected model. RESULTS: A total of 6,760 patients were included, of whom 2,762 (40.9%) filled at least 1 opioid prescription >90 days after surgery. The artificial neural network model showed superior discrimination and calibration with area under the receiver operating curve = 0.71 (95% confidence interval 0.68-0.74) and Brier score = 0.21 (95% confidence interval 0.20-0.22). Postsurgical opioid use, age, and preoperative opioid use had the most influence on model outcome. Lesser factors included the presence of a psychological comorbidity and strong history of a substance use disorder. CONCLUSIONS: The artificial neural network model shows sufficient validity and discrimination for use in clinical practice. The 5 identified factors (age, preoperative opioid use, postoperative opioid use, presence of a mental health comorbidity, and presence of a preoperative substance use disorder) accurately predict the likelihood of prolonged opioid use following hip arthroscopy. LEVEL OF EVIDENCE: III, retrospective comparative prognostic trial.

Can Machine-learning Algorithms Predict Early Revision TKA in the Danish Knee Arthroplasty Registry?

BACKGROUND: Revision TKA is a serious adverse event with substantial consequences for the patient. As the demand for TKA rises, reducing the risk of revision TKA is becoming increasingly important. Predictive tools based on machine-learning algorithms could reform clinical practice. Few attempts have been made to combine machine-learning algorithms with data from nationwide arthroplasty registries and, to the authors' knowledge, none have tried to predict the likelihood of early revision TKA. QUESTION/PURPOSES: We used the Danish Knee Arthroplasty Registry to build models to predict the likelihood of revision TKA within 2 years of primary TKA and asked: (1) Which preoperative factors were the most important features behind these models' predictions of revision? (2) Can a clinically meaningful model be built on the preoperative factors included in the Danish Knee Arthroplasty Registry? METHODS: The Danish Knee Arthroplasty Registry collects patients' characteristics and surgical information from all arthroplasties conducted in Denmark and thus provides a large nationwide cohort of patients undergoing TKA. As training dataset, we retrieved all preoperative variables of 25,104 primary TKAs from 2012 to 2015. The same variables were retrieved from 6170 TKAs conducted in 2016, which were used as a hold-out year for temporal external validation. If a patient received bilateral TKA, only the first knee to receive surgery was included. All patients were followed for 2 years, with removal, exchange, or addition of an implant defined as TKA revision. We created four different predictive models to find the best performing model, including a regression-based model using logistic regression with least shrinkage and selection operator (LASSO), two classification tree models (random forest and gradient boosting model) and a supervised neural network. For comparison, we created a noninformative model predicting that all observations were unrevised. The four machine learning models were trained using 10-fold cross-validation on the training dataset after adjusting for the low percentage of revisions by over-sampling revised observations and undersampling unrevised observations. In the validation dataset, the models' performance was evaluated and compared by density plot, calibration plot, accuracy, Brier score, receiver operator characteristic (ROC) curve and area under the curve (AUC). The density plot depicts the distribution of probabilities and the calibration plot graphically depicts whether the predicted probability resembled the observed probability. The accuracy indicates how often the models' predictions were correct and the Brier score is the mean distance from the predicted probability to the observed outcome. The ROC curve is a graphical output of the models' sensitivity and specificity from which the AUC is calculated. The AUC can be interpreted as the likelihood that a model correctly classified an observation and thus, a priori, an AUC of 0.7 was chosen as threshold for a clinically meaningful model. RESULTS: Based the model training, age, postfracture osteoarthritis and weight were deemed as important preoperative factors within the machine learning models. During validation, the models' performance was not different from the noninformative models, and with AUCs ranging from 0.57 to 0.60, no models reached the predetermined AUC threshold for a clinical useful discriminative capacity. CONCLUSION: Although several well-known presurgical risk factors for revision were coupled with four different machine learning methods, we could not develop a clinically useful model capable of predicting early TKA revisions in the Danish Knee Arthroplasty Registry based on preoperative data. CLINICAL RELEVANCE: The inability to predict early TKA revision highlights that predicting revision based on preoperative information alone is difficult. Future models might benefit from including medical comorbidities and an anonymous surgeon identifier variable or may attempt to build a postoperative predictive model including intra- and postoperative factors as these may have a stronger association with early TKA revisions.

Can Predictive Modeling Tools Identify Patients at High Risk of Prolonged Opioid Use After ACL Reconstruction?

BACKGROUND: Machine-learning methods such as the Bayesian belief network, random forest, gradient boosting machine, and decision trees have been used to develop decision-support tools in other clinical settings. Opioid abuse is a problem among civilians and military service members, and it is difficult to anticipate which patients are at risk for prolonged opioid use. QUESTIONS/PURPOSES: (1) To build a cross-validated model that predicts risk of prolonged opioid use after a specific orthopaedic procedure (ACL reconstruction), (2) To describe the relationships between prognostic and outcome variables, and (3) To determine the clinical utility of a predictive model using a decision curve analysis (as measured by our predictive system's ability to effectively identify high-risk patients and allow for preventative measures to be taken to ensure a successful procedure process). METHODS: We used the Military Analysis and Reporting Tool (M2) to search the Military Health System Data Repository for all patients undergoing arthroscopically assisted ACL reconstruction (Current Procedure Terminology code 29888) from January 2012 through December 2015 with a minimum of 90 days postoperative follow-up. In total, 10,919 patients met the inclusion criteria, most of whom were young men on active duty. We obtained complete opioid prescription filling histories from the Military Health System Data Repository's pharmacy records. We extracted data including patient demographics, military characteristics, and pharmacy data. A total of 3.3% of the data was missing. To curate and impute all missing variables, we used a random forest algorithm. We shuffled and split the data into 80% training and 20% hold-out sets, balanced by outcome variable (Outcome90Days). Next, the training set was further split into training and validation sets. Each model was built on the training data set, tuned with the validation set as applicable, and finally tested on the separate hold-out dataset. We chose four predictive models to develop, at the end choosing the best-fit model for implementation. Logistic regression, random forest, Bayesian belief network, and gradient boosting machine models were the four chosen models based on type of analysis (classification). Each were trained to estimate the likelihood of prolonged opioid use, defined as any opioid prescription filled more than 90 days after anterior cruciate reconstruction. After this, we tested the models on our holdout set and performed an area under the curve analysis concordance statistic, calculated the Brier score, and performed a decision curve analysis for validation. Then, we chose the method that produced the most suitable analysis results and, consequently, predictive power across the three calculations. Based on the calculations, the gradient boosting machine model was selected for future implementation. We systematically selected features and tuned the gradient boosting machine to produce a working predictive model. We performed area under the curve, Brier, and decision curve analysis calculations for the final model to test its viability and gain an understanding of whether it is possible to predict prolonged opioid use. RESULTS: Four predictive models were successfully developed using gradient boosting machine, logistic regression, Bayesian belief network, and random forest methods. After applying the Boruta algorithm for feature selection based on a 100-tree random forest algorithm, features were narrowed to a final seven features. The most influential features with a positive association with prolonged opioid use are preoperative morphine equivalents (yes), particular pharmacy ordering sites locations, shorter deployment time, and younger age. Those observed to have a negative association with prolonged opioid use are particular pharmacy ordering sites locations, preoperative morphine equivalents (no), longer deployment, race (American Indian or Alaskan native) and rank (junior enlisted).On internal validation, the models showed accuracy for predicting prolonged opioid use with AUC greater than our benchmark cutoff 0.70; random forest were 0.76 (95% confidence interval 0.73 to 0.79), 0.76 (95% CI 0.73 to 0.78), 0.73 (95% CI 0.71 to 0.76), and 0.72 (95% CI 0.69 to 0.75), respectively. Although the results from logistic regression and gradient boosting machines were very similar, only one model can be used in implementation. Based on our calculation of the Brier score, area under the curve, and decision curve analysis, we chose the gradient boosting machine as the final model. After selecting features and tuning the chosen gradient boosting machine, we saw an incremental improvement in our implementation model; the final model is accurate, with a Brier score of 0.10 (95% CI 0.09 to 0.11) and area under the curve of 0.77 (95% CI 0.75 to 0.80). It also shows the best clinical utility in a decision curve analysis. CONCLUSIONS: These scores support our claim that it is possible to predict which patients are at risk of prolonged opioid use, as seen by the appropriate range of hold-out analysis calculations. Current opioid guidelines recommend preoperative identification of at-risk patients, but available tools for this purpose are crude, largely focusing on identifying the presence (but not relative contributions) of various risk factors and screening for depression. The power of this model is that it will permit the development of a true clinical decision-support tool, which risk-stratifies individual patients with a single numerical score that is easily understandable to both patient and surgeon. Probabilistic models provide insight into how clinical factors are conditionally related. Not only will this gradient boosting machine be used to help understand factors contributing to opiate misuse after ACL reconstruction, but also it will allow orthopaedic surgeons to identify at-risk patients before surgery and offer increased support and monitoring to prevent opioid abuse and dependency. LEVEL OF EVIDENCE: Level III, therapeutic study.

RNA expression in human retina.

Recent Genome-wide Association Studies (GWASs) for eye diseases/traits have delivered a number of novel findings across a diverse range of diseases, including age-related macular degeneration (AMD), glaucoma and refractive error. However, despite this astonishing rate of success, the major challenge still remains to not only confirm that the genes implicated in these studies are truly the genes conferring protection from or risk of disease but also to define the functional roles these genes play in disease. Ongoing evidence is accumulating that the single nucleotide polymorphisms (SNPs) used in GWAS and fine mapping studies have causal effects through their influence on gene expression rather than affecting protein function. The biological interpretation of SNP regulatory effects for a tissue requires knowledge of the transcriptome for that tissue. We summarize the reasons to characterize the complete retinal transcriptome as well as the evidence to include an assessment of differences in regional retinal expression.

Utilization of the Department of Defense Peer-Reviewed Orthopaedic Research Program (PRORP): Combating Musculoskeletal Disease With PRORP.

INTRODUCTION: Established in 2009, the Department of Defense (DoD) Peer-Reviewed Orthopaedic Research Program (PRORP) is an annual funding program for orthopaedic research that seeks to develop evidence for new clinical practice guidelines, procedures, technologies, and drugs. The aim was to help reduce the burden of injury for wounded Service members, Veterans, and civilians and to increase return-to-duty and return-to-work rates. Relative to its burden of disease, musculoskeletal injuries (MSKIs) are one of the most disproportionately underfunded conditions. The focus of the PRORP includes a broad spectrum of MSKI in areas related to unique aspect of combat- and some noncombat-related injuries. The PRORP may serve as an important avenue of research for nonmilitary communities by offering areas of shared interests for the advancement of military and civilian patient cohort MSKI care. The purpose of this study was to provide a descriptive analysis of the DoD PRORP, which is an underrecognized but high value source of research funding for a broad spectrum of both combat- and noncombat-related MSKIs. METHODS: The complete PRORP Funding Portfolio for FY2009-FY2017 was obtained from the Congressionally Directed Medical Research Programs (CDMRP), which includes 255 awarded grants. Information pulled from the CDMRP included awardee descriptors (sex, education level, affiliated institution type, research specialty, and previous award winner [yes/no]) and grant award descriptors (grant amount, year, primary and secondary awarded topics, research type awarded, and mechanism of award). Distribution statistics were broken down by principal investigator specialty, sex, degree, organization type, research type, mechanism, and research topics. Distribution and statistical analysis was applied using R software version 3.6.3. RESULTS: From FY2009 to 2017, $285 million was allocated for 255 PRORP-funded research studies. The seven major orthopaedic subspecialties (foot and ankle, hand, musculoskeletal oncology, pediatrics, spine, sports medicine, and trauma) were represented. Trauma and hand subspecialists received the largest amount of funding, approximately $28 (9.6%) and $22 million (7.1%), respectively. However, only 22 (8.6%) and 26 (10.2%) of the primary investigators were trauma and hand subspecialists, respectively. The primary research categories were diverse with the top five funded PRORP topics being rehabilitation ($53 million), consortia ($39 million), surgery ($37 million), device development ($30 million), and pharmacology ($10 million). DISCUSSION: The CDMRP funding represents an excellent resource for orthopaedic medical research support that includes trauma and nontrauma orthopaedic conditions. This study serves to promote and communicate the missions of the PRORP both within and beyond the DoD to raise awareness and expand access of available funding for orthopaedic focused research. SIGNIFICANCE/CLINICAL RELEVANCE: A likelihood exists that this project will provide sustained and powerful influence on future research by promoting awareness of orthopaedic funding sources. LEVEL OF EVIDENCE: Level III.

Development and comparison of 1-year survival models in patients with primary bone sarcomas: External validation of a Bayesian belief network model and creation and external validation of a new gradient boosting machine model.

BACKGROUND: Bone sarcomas often present late with advanced stage at diagnosis and an according, varying short-term survival. In 2016, Nandra et al. generated a Bayesian belief network model for 1-year survival in patients with bone sarcomas. The purpose of this study is: (1) to externally validate the prior 1-year Bayesian belief network prediction model for survival in patients with bone sarcomas and (2) to develop a gradient boosting machine model using Nandra et al.'s cohort and evaluate whether the gradient boosting machine model outperforms the Bayesian belief network model when externally validated in an independent Danish population cohort. MATERIAL AND METHODS: The training cohort comprised 3493 patients newly diagnosed with bone sarcoma from the institutional prospectively maintained database at the Royal Orthopaedic Hospital, Birmingham, UK. The validation cohort comprised 771 patients with newly diagnosed bone sarcoma included from the Danish Sarcoma Registry during January 1, 2000-June 22, 2016. We performed area under receiver operator characteristic curve analysis, Brier score and decision curve analysis to evaluate the predictive performance of the models. RESULTS: External validation of the Bayesian belief network 1-year prediction model demonstrated an area under receiver operator characteristic curve of 68% (95% confidence interval, 62%-73%). Area under receiver operator characteristic curve of the gradient boosting machine model demonstrated: 75% (95% confidence interval: 70%-80%), overall model performance by the Brier score was 0.09 (95% confidence interval: 0.077-0.11) and decision curve analysis demonstrated a positive net benefit for threshold probabilities above 0.5. External validation of the developed gradient boosting machine model demonstrated an area under receiver operator characteristic curve of 63% (95% confidence interval: 57%-68%), and the Brier score was 0.14 (95% confidence interval: 0.12-0.16). CONCLUSION: External validation of the 1-year Bayesian belief network survival model yielded a poor outcome based on a Danish population cohort validation. We successfully developed a gradient boosting machine 1-year survival model. The gradient boosting machine did not outperform the Bayesian belief network model based on external validation in a Danish population-based cohort.

Genome-wide association meta-analysis highlights light-induced signaling as a driver for refractive error.

Refractive errors, including myopia, are the most frequent eye disorders worldwide and an increasingly common cause of blindness. This genome-wide association meta-analysis in 160,420 participants and replication in 95,505 participants increased the number of established independent signals from 37 to 161 and showed high genetic correlation between Europeans and Asians (>0.78). Expression experiments and comprehensive in silico analyses identified retinal cell physiology and light processing as prominent mechanisms, and also identified functional contributions to refractive-error development in all cell types of the neurosensory retina, retinal pigment epithelium, vascular endothelium and extracellular matrix. Newly identified genes implicate novel mechanisms such as rod-and-cone bipolar synaptic neurotransmission, anterior-segment morphology and angiogenesis. Thirty-one loci resided in or near regions transcribing small RNAs, thus suggesting a role for post-transcriptional regulation. Our results support the notion that refractive errors are caused by a light-dependent retina-to-sclera signaling cascade and delineate potential pathobiological molecular drivers.