Machine Learning Algorithm Helps Identify Non-Diagnosed Prodromal Alzheimer's Disease Patients in the General Population

BACKGROUND

Recruiting patients for clinical trials of potential therapies for Alzheimer's disease (AD) remains a major challenge, with demand for trial participants at an all-time high. The AD treatment R and D pipeline includes around 112 agents. In the United States alone, 150 clinical trials are seeking 70,000 participants. Most people with early cognitive impairment consult primary care providers, who may lack time, diagnostic skills and awareness of local clinical trials. Machine learning and predictive analytics offer promise to boost enrollment by predicting which patients have prodromal AD, and which will go on to develop AD.

OBJECTIVES

The authors set out to develop a machine learning predictive model that identifies prodromal AD patients in the general population, to aid early AD detection by primary care physicians and timely referral to expert sites for biomarker confirmation of diagnosis and clinical trial enrollment.

DESIGN

The authors use a classification machine learning algorithm to extract patterns within healthcare claims and prescription data three years prior to AD diagnosis/AD drug initiation.

SETTING

The study focused on subjects included within proprietary IQVIA US data assets (claims and prescription databases). Patient information was extracted from January 2010 to July 2018, for cohorts aged between 50 and 85 years.

PARTICIPANTS

A total of 88,298,289 subjects aged between 50 and 85 years were identified. For the positive cohort, 667,288 subjects were identified who had 24 months of medical history and at least one record with AD or AD treatment. For the negative cohort, 3,670,254 patients were selected who had a similar length of medical history and who were matched to positive cohort subjects based on the prevalence rate. The scoring cohort was selected based on availability of recent medical data of 2-5 years and included 72,670,283 subjects between the ages of 50 and 85 years. Intervention (if any): None.

MEASUREMENTS

A list of clinically-relevant and interpretable predictors was generated and extracted from the data sets for each subject, including pharmacological treatments (NDC/product), office/specialist visits (specialty), tests and procedures (HCPCS and CPT), and diagnosis (ICD). The positive cohort was defined as patients who have AD diagnosis/AD treatment with a 3 years offset as an estimate for prodromal AD diagnosis. Supervised ML techniques were used to develop algorithms to predict the occurrence of prodromal AD cases. The sample dataset was divided randomly into a training dataset and a test dataset. The classification models were trained and executed in the PySpark framework. Training and evaluation of LogisticRegression, DecisionTreeClassifier, RandomForestClassifier, and GBTClassifier were executed using PySpark's mllib module. The area under the precision-recall curve (AUCPR) was used to compare the results of the various models.

RESULTS

The AUCPRs are 0.426, 0.157, 0.436, and 0.440 for LogisticRegression, DecisionTreeClassifier, RandomForestClassifier, and GBTClassifier, respectively, meaning that GBTClassifier (Gradient Boosted Tree) outperforms the other three classifiers. The GBT model identified 222,721 subjects in the prodromal AD stage with 80% precision. Some 76% of identified prodromal AD patients were in the primary care setting.

CONCLUSIONS

Applying the developed predictive model to 72,670,283 U.S. residents, 222,721 prodromal AD patients were identified, the majority of whom were in the primary care setting. This could drive major advances in AD research by enabling more accurate and earlier prodromal AD diagnosis at the primary care physician level , which would facilitate timely referral to expert sites for in-depth assessment and potential enrolment in clinical trials.

Endogenous glucocorticoids attenuate Shiga toxin-2-induced toxicity in a mouse model of haemolytic uraemic syndrome

The concept that during an immune challenge the release of glucocorticoids (GC) provides feedback inhibition on evolving immune responses has been drawn primarily from studies of autoimmune and/or inflammatory processes in animal models. The epidemic form of haemolytic uraemic syndrome (HUS) occurs secondary to infection with Gram-negative bacteria that produce Shiga toxin (Stx). Although Stx binding to the specific receptors present on renal tissue is the primary pathogenic mechanism, inflammatory or immune interactions are necessary for the development of the complete form of HUS. The aim of this study was to investigate the influence of endogenous GC on Stx-toxicity in a mouse model. Stx2 was injected into GC-deprived mice and survival rate, renal damage and serum urea levels were evaluated. Plasma corticosterone and cytosolic GC receptor (GR) concentration were also determined at multiple intervals post-Stx2 treatment. Higher sensitivity to Stx2 was observed in mice lacking endogenous GC, evidenced by an increase in mortality rates, circulating urea levels and renal histological damage. Moreover, Stx2 injection was associated with a transient but significant rise in corticosterone secretion. Interestingly, 24 h after Stx inoculation significant increases in total GR were detected in circulating neutrophils. These results indicate that interactions between the neuroendocrine and immune systems can modulate the level of damage significantly during a bacterial infection.

Fibrinogen promotes neutrophil activation and delays apoptosis

The acute phase of the inflammatory response involves an increase in the concentrations of different plasma proteins that include fibrinogen (Fbg) and multiple proinflammatory mediators. In parallel, neutrophil activation is thought to play a crucial role in several inflammatory conditions, and it has been recently demonstrated that Fbg specifically binds to the alpha-subunit of CD11b/CD18 on neutrophil surface. Although several reports have shown that CD11b engagement modulates neutrophil responses, the effect of human Fbg (hFbg), one of CD11b physiologic ligands, has not been exhaustively investigated. We have now shown that incubation of purified neutrophils with hFbg induces a transient and rapid elevation of free intracellular Ca2+. This early intracellular signal is accompanied by changes in the expression of neutrophil activation markers, including enhancement of CD11b and CD66b, and down-regulation of FcgammaRIII. In addition, we have evaluated the effect of hFbg on two functional events related to expression and resolution of inflammation: cytotoxic capacity and rate of neutrophil apoptosis. We have found that activation of neutrophils by hFbg resulted in both enhancement of phagocytosis and Ab-dependent cellular cytotoxicity, and delay of apoptosis. We conclude that during inflammatory processes, soluble Fbg could influence neutrophil responses, increasing and prolonging their functional capacity.

[The treatment of iatrogenic spurious aneurysm of the femoral artery by direct thrombin injection]

BACKGROUND AND OBJECTIVE

After percutaneous catheter introduction a false aneurysm occasionally develops at the site of puncture. This has been treated either surgically or, more recently, by ultrasound-guided compression. A new method has been tried in which the false aneurysm is thrombosed by injecting thrombin into it.

PATIENTS AND METHODS

In 29 patients thrombin was injected directly into the false aneurysm of the femoral artery, caused by catheter introduction into the vessel. Puncture of the aneurysm and injection of the thrombin solution was performed with continuous duplex-sonographic monitoring. The patients' age ranged from 42 to 88 years (mean 71 +/- 12 years). The false aneurysm had occurred after diagnostic catheterization (n = 5), balloon dilatation of peripheral vessels (n = 5) or balloon catheter dilatation of the coronary arteries (n = 19) with catheters size 5 F (n = 4), 6 F (n = 6), 8 F (n = 16) or 9-13 F (n = 3). The catheterization had been done 1-30 days previously (mean 5.3 +/- 6.9 days). The diameter of the aneurysm ranged from 2.1 to 5 cm (mean 3.5 +/- 0.9 cm).

RESULTS

The aneurysms thrombosed within seconds after injection of 0.075 to 1.5 ml (mean 0.4 +/- 0.4 ml). All interventions were successful and without complications. Any resulting haematoma regressed within a few days to a few weeks and none recurred. In two patients a persisting haematoma had later to be removed surgically, and in another patient a second aneurysm was removed surgically without prior thrombin injection.

CONCLUSION

A false aneurysm of the femoral artery caused by percutaneous catheterization can be successfully thrombosed by direct thrombin injection.

Shiga toxin-2 induces neutrophilia and neutrophil activation in a murine model of hemolytic uremic syndrome

It has been demonstrated that infections due to Shiga toxins (Stx) producing Escherichia coli are the main cause of the hemolytic uremic syndrome (HUS). Although it is recognized that Stx damage the glomerular endothelium, clinical and experimental evidence suggests that the inflammatory response is able to potentiate Stx toxicity. Lipopolysaccharides (LPS) and neutrophils (PMN) represent two central components of inflammation during a gram-negative infection. In this regard, patients with high peripheral PMN counts at presentation have a poor prognosis. Since the murine model has been used to study LPS-Stx interactions, we analyzed the effects of Stx alone or in combination with LPS on the kinetics of neutrophil production and activation and their participation in renal damage. We observed a sustained neutrophilia after Stx2 injection. Moreover, these neutrophils showed increased expression of CD11b, enhanced cytotoxic capacity, and greater adhesive properties. Regarding the cooperative effects of LPS on Stx2 action, we demonstrated potentiation of neutrophilia and CD11b induction at early times by pretreatment with LPS. Finally, a positive correlation between neutrophil percentage and renal damage (assayed as plasmatic urea) firmly suggests a role for PMN in the pathogenesis of HUS.

Pretreatment of mice with lipopolysaccharide (LPS) or IL-1beta exerts dose-dependent opposite effects on Shiga toxin-2 lethality

Haemolytic uraemic syndrome (HUS) has been closely associated with infection with a group of Shiga toxin-producing enterohaemorrhagic Eschericchia coli in young children. Shiga toxins (Stx) have been implicated as pathogenic agents of HUS by binding to the surface receptor of endothelial cells. LPS is a central product of the Gram-negative bacteria and several reports have documented that both LPS and Stx are important for disease development. In this study the reciprocal interactions between LPS and Stx2 are analysed in a mouse model. The results demonstrated that LPS was able to reduce or enhance Stx2 toxicity, depending on the dose and the timing of the injection. The involvement of the main early cytokines induced by LPS, tumour necrosis factor alpha (TNF-alpha) and IL-1beta, in those LPS opposite effects on Stx2 toxicity was evaluated. Stx2 toxicity was enhanced by in vivo injection of murine TNF-alpha and low doses of murine IL-1beta. However, at higher doses of IL-1beta which induced corticosteroid increase in serum, Stx2 lethality was decreased. Considering that dexamethasone and IL-1beta reproduce the LPS protective effects, it is suggested that endogenous corticosteroids secondary to the inflammatory response induced by LPS, mediate the protection against Stx2. It can be concluded that the fine equilibrium between proinflammatory and anti-inflammatory activities strongly influences Stx2 toxicity.

Lipopolysaccharide enhances FcgammaR-dependent functions in vivo through CD11b/CD18 up-regulation

Fc receptors for immunoglobulin G (IgG) (FcgammaR) mediate several defence mechanisms in the course of inflammatory and infectious diseases. In Gram-negative infections, cellular wall lipopolysaccharides (LPS) modulate different immune responses. We have recently demonstrated that murine LPS in vivo treatment significantly increases FcgammaR-dependent clearance of immune complexes (IC). In addition, we and others have reported the induction of adhesion molecules on macrophages and neutrophils by LPS in vivo and by tumour necrosis factor-alpha (TNF-alpha) in vitro. The aim of this paper was to investigate CD11b/CD18 participation in LPS enhancing effects on Fcgamma-dependent functionality of tissue macrophages. Our results have demonstrated that LPS can enhance antibody-dependent cellular cytotoxicity (ADCC) and IC-triggered cytotoxicity (IC-Ctx), two reactions which involve the Fcgamma-receptor but different lytic mechanisms. In vitro incubation of splenocytes from LPS-treated mice with anti-CD11b/CD18 abrogated ADCC and IC-Ctx enhancement, without affecting FcgammaR expression. Similar results were obtained with physiological concentrations of fibrinogen. In this way cytotoxic values of LPS-splenocytes decreased to the basal levels of control mice. Time and temperature requirements for such inhibition strongly suggested that anti-CD11b/CD18 could modulate intracellular signals leading to downregulation of FcgammaR functionality. Data presented herein support the hypothesis that functional and/or physical associations between integrins and FcgammaR could be critical for the modulation of effector functions during an inflammatory response.

Salicylic Acid Is Needed in Hypersensitive Cell Death in Soybean but Does Not Act as a Catalase Inhibitor

The function of salicylic acid (SA) in hypersensitive cell death was studied in a soybean (Glycine max)-Pseudomonas syringae pv glycinea system. The infection of cell cultures with bacteria leads to a hypersensitive reaction (HR), which is dependent on an appropriate avirulence gene and on low concentrations of SA. The requirement for SA is essential for a process shortly before the onset of the HR-caused cell death 5 to 6 h after infection with bacteria. SA given 10 to 12 h after infection or preincubation cannot rescue the completion of the cell death program. SA does not inhibit catalase or ascorbate peroxidase in soybean. In addition, the in vivo capacity of the cell culture for the rapid metabolism of H2O2 is not altered by SA. This clearly shows that SA is needed for the HR-caused cell death for a reaction downstream of the oxidative burst. Lipid peroxides accumulate during the HR, but the loss of membrane control precedes the generation of lipid peroxides. The accumulation of lipid peroxides in the HR can be prevented by lipid antioxidants. Nevertheless, cell death kinetics remain unaltered in the presence of antioxidants. It is concluded that lipid peroxides are a consequence of cell death, but not the primary cause of it.

Brachial artery approach for transluminal angioplasty of the internal carotid artery

One possible problem in internal carotid angioplasty is inaccessibility of the lesion due to elongation of the aortic arch, the brachiocephalic trunk, or the carotid artery itself. A new approach to performing angioplasty of the right or left internal carotid artery utilizing the brachial artery was used after failure of the transfemoral approach in 5 lesions (4 patients). The common carotid artery was cannulated with performed 5F catheters. Angioplasty was performed with a conventional balloon dilatation catheter. If required, a Wall stent was implanted to optimize the angiographic result. After failure of the conventional transfemoral technique, the brachial technique permitted successful angioplasty of the ipsilateral internal carotid artery in 4 lesions and the contralateral internal carotid artery in 1 lesion. In 1 patient, a stent was implanted. No complications occurred. The mean stenosis diameter decreased from 77.8 +/- 6.3% to 17.8 +/- 9.1%. Doppler sonography performed 4-6 months later showed no restenosis. The brachial artery approach seems to be a suitable alternative to the femoral technique.

[-Occlusion of atrial septal defect with a new occlusive device-]

An atrial septum defect was closed with an ASDOS (Babic) occluder in 13 patients aged from 22 to 67 years. The diameters of the ASD ranged from 12 to 36 mm, the left to right shunt from 35% to 70%. With one exception, the occluder could be implanted in all patients. Two patients with an oversized ASD (diameter 31 and 36 mm respectively) had to be operated after 8 h and 2 weeks, respectively. In one patient, a small asymptomatic hemopericardium was detected after 24 h by routine echocardiogram and in another patient a single umbrella arm fracture was noticed by routine x-ray 4 months after the implantation. No further complications occurred. Follow-up is now 3 months to 1 year. With one exception, there was no residual shunt as measured by oximetry.