Diagnostic Applications of Microarrays and Gene Expression in Transplants and Native Organs

• Author / Creator

  Madill-Thomsen, Katelynn S.

• There is a major unmet need for improved accuracy and precision in the diagnoses of transplant rejection and diseases causing tissue injury. Diagnoses relying on histologic assessments and visual assessments demonstrate significant variation between expert observers (as represented by low kappa values) and cannot assess many biological processes because they do not produce histologic changes. Arbitrarily assigned rules determined by consensus may or may not reflect the true disease phenotype, and the lack of objective diagnostic information presents a challenge to the clinician who is managing the patient’s care and making therapeutic decisions. Risks of over- or under-treatment can be serious: many therapies for transplant rejection or for primary diseases are expensive and carry risk for significant adverse effects. Improved diagnostic methods could alleviate healthcare costs by preventing treatment errors, increase treatment efficacy, and ultimately improve outcomes. Molecular diagnostic assessments using microarrays combined with machine learning algorithms for interpretation have shown promise for increasing diagnostic precision via probabilistic assessments, recalibrating standard-of-care (SOC) diagnostic methods, and clarifying ambiguous cases. This approach can use ensembles of algorithms to increase stability and can provide novel mechanistic insights. These features would benefit biopsy-dependent areas of medicine like transplantation or management of inflammatory diseases such as ulcerative colitis (UC). The analyses described in this thesis are based on the hypothesis that new molecular systems for biopsy interpretation (i.e. the Molecular Microscope Diagnostic System ‘MMDx’) would provide insights on disease processes and highly reproducible results from a comparatively small amount of tissue, and would constitute a general approach that could be useful in many new areas of medicine; both in transplantation and in diseases in native organs.Analyses first focused on establishing the reproducibility and robustness of the techniques used in MMDx, and its relationship to SOC approaches currently in use. The effects of tissue heterogeneity on the MMDx output was studied using kidney transplant cortex and medulla biopsy samples. The frequency and pattern of discrepancies within MMDx-Kidney (between expert observers), within histology (comparing the SOC diagnosis to a diagnosis assigned by an algorithm strictly following Banff guidelines), and finally between MMDx and histology were studied. Once the MMDx-Kidney test was well-defined, the MMDx system was translated into liver transplants where biopsy assessment is more challenging. MMDx-Liver was assessed with regards to its performance for diagnoses of both T cell-mediated rejection (TCMR) and various forms of injury. Although antibody-mediated rejection (ABMR) remains a controversial diagnosis in liver transplantation, previously annotated rejection-associated transcripts (RATs) were used to search the liver biopsy population for an ABMR phenotype analogous to that in kidney and heart. TCMR in liver biopsies was defined using archetypal analysis (AA), and the molecular findings were compared to SOC histology features. Injury was described in a liver biopsy population also using AA, and a classifier was developed for estimating histologic steatohepatitis (AUC=0.84). Finally, an MMDx system was developed for biopsies of a native organ disease: colon biopsies from patients with diagnosed UC. This was done to assess if methods developed for transplant biopsies have the potential for diagnostics of primary diseases in native tissue. These analyses found that there are multiple immune processes involved in UC disease activity: a dominant inflammatory process mediated by innate immunity and an underlying subtle T cell process. UC was examined using AA, revealing heterogeneity in the biopsy population that was not related to the SOC Endoscopic Mayo Score. This finding suggested that the SOC is not capturing all the information describing disease activity.This thesis explores a diagnostic system that would fulfill unmet needs in transplanted organs (kidney, liver) as well as in native organs (UC). Main findings indicate that MMDx has major implications for kidney transplant biopsies but can also be expanded and translated for use in transplanted livers (see Chapters 6 and 7). The MMDx approach has translational potential in understanding native organ biopsies, for the diagnosis of disease, tissue injury, and loss of function e.g. UC. If molecular diagnostic systems e.g. MMDx are responsibly developed with proper statistical and machine-learning techniques, appropriately validated, and well-defined in terms of observer guidelines for diagnoses, they have the potential to address the current unmet needs for precise and accurate assessments that clinicians are requesting to improve patient care.

• Subjects / Keywords

  • transplant rejection
  • transplantation
  • molecular diagnostics
  • gene expression
  • transplant injury
  • microarrays

• Graduation date

  Fall 2020

• Type of Item

  Thesis

• Degree

  Doctor of Philosophy

• DOI

  https://doi.org/10.7939/r3-gc9w-3587

• License

  Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.