English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Background During the last twenty years many tools, based on mathematical models, have been developed to predict ESKD at the time of kidney biopsy in patients with IgAN. The main limitation of these tools is the time frame to reach the ESKD. Recently, we have developed a Clinical Decision Support System (CDSS) (KI 2020) which includes 6 variables at the time of kidney biopsy: age, sex, hypertension, serum creatinine, daily proteinuria and score of the renal lesions according to the MEST-C classification. The tool (www.igan.poliba.it) is based on two different artificial neural networks of which the first predicts ESKD and the second one predicts the time to reach this outcome. Aim of our study has been to analyze the causes of discordance to predict or not ESKD in a cohort of 1116 IgAN patients (VALIGA cohort and external cohort) with a median follow-up of 88 months (49-135). To our knowledge this is the first report which analyzes the discrepant results of a toll to predict ESKD in IgAN patients. Method Demographic and laboratory data have been analyzed using means and standard deviations (SD) for continuous variables normally distributed, median in presence of non-normally distribution. Categorical variables are expressed as percentages. The means have been compared by the Student’s test and the medians using the Mann-Whitney U test. All the data were collected and stored in a standard Excel database. The statistical significance value p&amp;lt; 0.05 has been adopted. Results Discordance to reach or not ESKD was found in 216 patients (19.4%). In 77 patients with no prediction of ESKD, 25 patients with proteinuria &amp;gt; 0.5 g/day did not receive therapy after kidney biopsy or were cared very late. Failure of therapy (RASBs alone in 44 subjects and corticosteroids in 8 individuals) was observed in 52 patients. In 139 patients who did not reach ESKD but our tool predicted this outcome, interestingly, we found that 106 proteinuric patients (22 with GFR &amp;gt;50 ml/min/1.73 m2 and 84 with GFR &amp;lt; 50 ml/min/1.73 m2) had an improvement of the clinical course (reduction of proteinuria and stabilization of GFR value) after corticosteroid therapy. The remaining 33 patients benefited of RASB therapy. Among 84 patients with GFR&amp;lt; 50 ml/min/1.73 m2 36 had nephrotic proteinuria and were responsive to corticosteroids. These results suggest that we cannot rule out corticosteroid therapy in patients with reduced GFR and proteinuria &amp;gt; 3 g/day. Furthermore, our tool predicts the time frame to reach ESKD and indicates the potential effect of some drugs (RASBs, corticosteroids or their combination) to delay the outcome. We observed that corticosteroids combined with RASBs delayed more than 10 years the time to reach ESKD. Conclusions Our tool predicted ESKD in a percentage higher than that observed but, interestingly, it was found that a high number of patients benefited of corticosteroids in combination with RASBs. Furthermore, our tool predicted time to reach ESKD and indicated the potential benefit of therapy to delay the crude outcome. In fact, therapy delayed the outcome of more than 10 years when combination of corticosteroids and RASBs was administered. Therefore, our tool shows to physicians that their patients may delay the ESKD receiving corticosteroids alone or in combination with RASBs. This approach is important because in many cases it is a strengthen point to convince patients to accept the prescribed therapy. Moreover, this report shows for the first time that predicted ESKD may be delayed administering personalized therapy suggested by our tool. Essentially, in a high percentage of patients the failure of our tool is not an error but the positive effect of therapy or misconduct of patients management.

MO260PERFORMANCE ANALYSIS OF AN ARTIFICIAL NEURAL NETWORK TOOL TO PREDICT ESKD IN PATIENTS WITH IMMUNOGLOBULIN A NEPHROPATHY (IGAN)