Biološki procesi in napovedovanje neodzivnosti na zaviralce dejavnika tumorske nekroze pri Crohnovi bolezni z integracijo genomskih podatkov

Jezernik, Gregor

| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

First page > Show document

Show document

Title:	Biološki procesi in napovedovanje neodzivnosti na zaviralce dejavnika tumorske nekroze pri Crohnovi bolezni z integracijo genomskih podatkov
Authors:	ID Jezernik, Gregor (Author) ID Potočnik, Uroš (Mentor) More about this mentor... ID Skok, Pavel (Comentor)
Files:	DOK_Jezernik_Gregor_2021.pdf (3,69 MB) MD5: 0D0734701DF7AC391843FC938D37C812 PID: 20.500.12556/dkum/72840819-aafc-416a-9ddd-eb3d24bd498f
Language:	Slovenian
Work type:	Doctoral dissertation
Typology:	2.08 - Doctoral Dissertation
Organization:	MF - Faculty of Medicine
Abstract:	Razvoj bioloških zdravil je pomembno prispeval k možnostim zdravljenja raka in imunsko pogojenih bolezni. Med najpogosteje uporabljenimi biološkimi zdravili so zaviralci dejavnika tumorske nekroze (TNF). Crohnova bolezen je pogosta imunsko pogojena bolezen prebavil, ki se zdravi z zaviralci TNF. Kljub tarčnemu delovanju zaviralcev TNF del bolnikov s Crohnovo boleznijo žal ne doseže dobrega odziva na zaviralce TNF že ob uvedbi terapije ali pa sprva dober odziv na zaviralce TNF s časom izzveni. Neodzivnost na zaviralce TNF predstavlja pomeni izgubo nadzora nad pogosto hudim bolezenskim stanjem bolnika s Crohnovo boleznijo, ki je po nepotrebnem izpostavljen potencialno hudim neželenim stranskim učinkom bioloških zdravil, in tudi precejšnje finančno breme za zdravstveno blagajno. Ti razlogi utemeljujejo potrebo po napovedovanju odziva na biološka zdravila, po možnosti še pred uvedbo zdravljenja. V doktorski disertaciji smo celostno raziskali biološke označevalce odziva na zaviralce dejavnika tumorske nekroze na ravni DNA in RNA ter maščobnih kislin v vzorcih periferne venske krvi skupine slovenskih bolnikov s Crohnovo boleznijo, ki se je zdravila z adalimumabom. Rezultate teh analiz smo uporabili za oblikovanje novih napovednih modelov s pristopi strojnega učenja, t.i. metode podpornih vektorjev. Za namene iskanja vzročnih bioloških procesov, ki pogojujejo neodzivnost na zaviralce dejavnika tumorske nekroze, smo sistematsko preučili gensko ontologijo že objavljenih bioloških označevalcev odziva na zaviralce dejavnika tumorske nekroze v kronični vnetni črevesni bolezni. Za primerjalno analizo genske ontologije smo zbrali tudi biološke označevalce odziva v revmatoidnem artritisu. Ker je neodzivnost pogostejša med pediatričnimi bolniki, smo analizo genske ontologije razširili še na vzročne gene pediatričnih dednih oblik kronične vnetne črevesne bolezni in sindrome s klinično sliko, skladno s kronično vnetno črevesno boleznijo. Dodatno smo tudi poskusili ponoviti že objavljen napovedni model odziva na infliksimab, ki temelji na izražanju petih genov v črevesni sluznici. Rezultati genske ontologije že objavljenih označevalcev kažejo na povezavo med krvnimi lipoproteini in odzivom na zaviralce dejavnika tumorske nekroze pri kronični vnetni črevesni bolezni, kot tudi pri revmatoidnem artritisu. Na osnovi rezultatov genske ontologije pediatričnih dednih oblik kronične vnetne črevesne bolezni lahko sklepamo, da so zelo zgodnje pediatrične oblike z nastopom bolezni pred šestim letom starosti ločena genetska entiteta in je neodzivnost pogojena z drugimi procesi, npr. s primarno imunsko pomanjkljivostjo. Analiza bioloških podatkov na ravni DNA in RNA ter maščobnih kislin ni pokazala biološkega označevalca, ki bi dosegel statistično značilnost, kar odraža tudi analiza napovedne moči s pristopi strojnega učenja. Profili maščobnih kislin nimajo napovedne moči za določevanje odziva na zaviralce dejavnika tumorske nekroze, genomski in transkripromski podatki pa imajo le nizko napovedno moč. Napovedni model na osnovi že objavljenega modela izražanja genov v črevesni sluznici smo uspešno ponovili in prenesli na drugo učinkovino (adalimumab). Na osnovi izražanja štirih genov v vneti in nevneti črevesni sluznici je možno napovedati odziv na zaviralce dejavnika tumorske nekroze s natančnostjo do 100 %. Nato smo še analizirali diagnostično napovedno moč bioloških podatkov s vključitvijo bioloških podatkov zdravih prostovoljcev, ki so že bili na voljo. Napovedni model na osnovi dednega zapisa in profilov maščobnih kislin je z natančnostjo do 100 % ločil med zdravimi prostovoljci in bolniki s Crohnovo boleznijo.
Keywords:	Crohnova bolezen, genomika, transkriptomika, adalimumab, izid zdravljenja, genska ontologija
Place of publishing:	Maribor
Year of publishing:	2020
PID:	20.500.12556/DKUM-77986
COBISS.SI-ID:	47827971
NUK URN:	URN:SI:UM:DK:CTZVUSVW
Publication date in DKUM:	20.01.2021
Views:	1462
Downloads:	150
Metadata:
Categories:	MF
:	Copy citation

Average score:	(0 votes)
Your score:	Voting is allowed only for logged in users.
Share:

Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Licences

License:	CC BY-NC-SA 4.0, Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International

Link:	http://creativecommons.org/licenses/by-nc-sa/4.0/
Description:	A Creative Commons license that bans commercial use and requires the user to release any modified works under this license.
Licensing start date:	29.09.2020

Secondary language

Language:	English
Title:	Biological processes and predicting non-response to tumor necrosis factor inhibitors in Crohn's disease by integrating genomic data
Abstract:	The development of biological drugs has significantly contributed to therapeutic options in cancer and immune-mediated chronic diseases. Tumor necrosis factor (TNF) inhibitors are one of the most commonly used biological drugs. Crohn's disease is a common gastrointestinal immune disease that is treated with TNF inhibitors. Despite the targeted action of TNF inhibitors, some patients with Crohn's disease unfortunately do not respond to TNF inhibitors or lose response to TNF inhibitors over time. Non-response to TNF inhibitors usually represents a loss of control over severe symptoms, unnecessarily exposes Crohn's disease patients to potentially severe side effects of biological medicines and represents a significant burden on the health fund. These reasons justify the need to predict the response to biological drugs, preferably before initiating treatment. In this doctoral dissertation we thoroughly investigated the biological markers of response to tumor necrosis factor inhibitors on the level of DNA, RNA and fatty acids in samples of peripheral venous blood of a group of Slovenian patients with Crohn's disease treated with adalimumab. The results were used to design new predictive models with machine learning approaches, i.e. support vector machines. To investigate biological processes causal for non-response to tumor necrosis factor inhibitors, we systematically examined the genetic ontology of previously published biological markers of response to tumor necrosis factor inhibitors in inflammatory bowel disease. We expanded the search to biological response markers in rheumatoid arthritis for comparative gene ontology analysis. Since non-response is more common among pediatric patients, we also extended the gene ontology analysis to causal genes for pediatric hereditary forms of chronic inflammatory bowel disease and syndromes with clinical characteristics consistent with inflammatory bowel disease. In addition, we attempted to replicate a published predictive model of response to infliximab based on the expression of five genes in the intestinal mucosa. The results of the gene ontology analysis of previously reported response markers suggest an association between blood lipoproteins and the response to tumor necrosis factor inhibitors in inflammatory bowel disease as well as in rheumatoid arthritis. Based on the results of gene ontology analysis of pediatric inherited forms of inflammatory bowel disease, we conclude that very early pediatric forms with an onset before the age of six are a separate genetic entity and are their non-response is likely caused by other processes, e.g. primary immune deficiency. Analysis of biological data on the level of DNA, RNA and fatty acids did not reveal biological markers with statistical significance, which is also reflected by the analysis of predictive power through machine learning approaches. Fatty acid profiles have no predictive power to determine response to tumor necrosis factor inhibitors while genomic and transcriptomic data has low predictive power. The predictive model based on the previously published intestinal mucosa gene expression model was successfully replicated and applied to another biological drug (adalimumab). Based on the expression of four genes in the inflamed and non-inflamed intestinal mucosa, it is possible to predict the response to tumor necrosis factor inhibitors with an accuracy of up to 100 %. We then further analyzed the diagnostic predictive power of biological data by incorporating already available biological data of healthy volunteers. A predictive model based on genetic data and fatty acid profiles distinguished between healthy volunteers and patients with Crohn's disease with an accuracy of up to 100 %.
Keywords:	Crohn's disease, genomics, transcriptomics, adalimumab, treatment outcome, gene ontology

Comments

Leave comment

You must log in to leave a comment.

Comments (0)

0 - 0 / 0

There are no comments!

Back