=Paper=
{{Paper
|id=Vol-3804/paper3dc
|storemode=property
|title=Information models in healthcare: CAR T-cell therapy
|pdfUrl=https://ceur-ws.org/Vol-3804/paper3dc.pdf
|volume=Vol-3804
|authors=Veronika Kostrouchová
|dblpUrl=https://dblp.org/rec/conf/bir/Kostrouchova24
}}
==Information models in healthcare: CAR T-cell therapy==
https://ceur-ws.org/Vol-3804/paper3dc.pdf
Information models in healthcare: CAR T-cell therapy
Veronika Kostrouchová1
1 Prague University of Economics and Business, W. Churchill Sq. 1938/4, Prague 3, Czech Republic
Abstract
The evolution of CAR T-cell therapy in cancer treatment represents a significant advancement in
precision medicine, necessitating robust models for effective implementation. This article explores
various modeling approaches to CAR T-cell therapy, ranging from high-level conceptual frameworks to
detailed business process models. This work introduces three primary models: the Ontological Model,
the PURO Model, and the MMABP Model. Each model serves distinct purposes: the Ontological Model
provides a high-level framework of the domain, establishing foundational concepts and relationships.
The PURO Model offers a flexible, graphical ontology sketching tool that enriches the initial
conceptualization into detailed operational frameworks. Lastly, the MMABP Model delves into the
specifics of the business process, detailing stages and interactions within the treatment workflow.
These models collectively foster a comprehensive understanding—from theoretical underpinnings to
actionable insights—enhancing the management and efficacy of CAR T-cell therapy.
Keywords
CAR T-cell therapy, MMABP, process map, conceptual model, PURO, ontoUML1
1. Introduction
1.1. The goal
The goal of this work is to showcase different approaches to modeling a specific cancer treatment
process (CAR T-cell therapy) from the most general to a more detailed business process model
and then eventually leading to a particular application using a decision engine as a long-term
vision of where to take this initiative. The use of these models represents a progression from a
general to a more focused view. Each model serves a specific purpose and level of detail:
The Ontological Model provides a high-level, generalized view of the domain, outlining key
concepts and their relationships. It sets the broad context and foundational understanding. An
ontological model is a structured representation of knowledge in a specific domain. It defines the
concepts, entities, and relationships within that domain, providing a framework to understand
and analyze complex systems or processes.
The PURO model serves as a flexible, graphical ontology sketching tool that allows for an
initial high-level conceptualization, which can be elaborated into more detailed and
operationally significant ontological frameworks.
The MMABP (Methodology for Modeling and Analysis of Business Processes) Model offers a
more detailed view, specifically focusing on the process aspect. It breaks down the general
concepts into specific stages and steps, providing a clearer picture of the flow and interactions.
It provides a structured approach for analyzing and representing business processes and
involves identifying, documenting, and analyzing the sequences of activities within an
organization to achieve specific business outcomes.
BIR-WS 2024: BIR 2024 Workshops and Doctoral Consortium, 23rd International Conference on Perspectives in
Business Informatics Research (BIR 2024), September 11-13, 2024, Prague, Czech Rep.
kosv08@vse.cz (V. Kostrouchová)
0000-0001-8417-6721 (V. Kostrouchová)
© 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
This approach allows for a comprehensive understanding that starts from a broad theoretical
base and narrows down to specific, practical applications. The following research questions were
set to guide the study and focus the analysis: 1. What are the key concepts and relationships
within the domain of CAR T-cell therapy? 2. What are the detailed stages and interactions within
the CAR T-cell therapy workflow, and how does the MMABP model improve the understanding
and optimization of these processes? 3. What potential challenges and outcomes can be
identified in the CAR T-cell therapy process through comprehensive modeling, and how can
these insights improve patient management and clinical decision-making?
1.2. Background
CAR T-cell therapy represents a significant leap in cancer treatment, shifting towards precision
medicine and immunotherapy. It utilizes the body's immune system, specifically engineering T
cells, to target and destroy cancer cells. This method is part of a broader category of targeted
therapies, which include monoclonal antibodies and tyrosine kinase inhibitors, all designed to
interfere with specific cancer cell processes. Immunotherapies like checkpoint inhibitors and
cancer vaccines aim to boost the immune system's natural ability to fight cancer, marking a
transformative approach in oncology.
Modeling CAR T-cell therapy is crucial for several reasons. It allows for more detailed
understanding and representation of this complex treatment's mechanisms. Modeling aids in
identifying potential challenges, side effects, and outcomes of therapy, facilitating the
development of more effective strategies and management approaches. Furthermore, it supports
the communication of complex information to researchers, clinicians, and patients, enhancing
collaboration and informed decision-making in cancer treatment.
The following flow chart covers the major steps involved in CAR T-cell therapy, from initial
patient reception to follow-up. It starts with a relapse of the patient’s disease and proceeds with
a referral from the patient’s oncologist leading to a registration with a clinic that provides CAR
T-cell therapy. Eligibility is screened, and upon consent, an appointment is assigned. The
treatment phase includes cell collection through leukapheresis, cell manufacturing and
proliferation, and then quality checks before patient preparation for the infusion of CAR T-cells
(this step can include chemotherapy for some patients). Post-infusion, the patient is monitored,
followed by care and response assessment. The flow chart was created based on the treatment
process as outlined by the National Comprehensive Cancer Network[1].
Figure 1: Simple CAR T-cell Therapy Treatment flowchart.
This above treatment flow chart is obviously lacking crucial information. One of its main
disadvantages is that it does not consider the various other possible outcomes besides
successfully continuing the process. In clinical or treatment processes, particularly for complex
procedures like CAR T-cell therapy, it's essential to include potential contingencies and
alternative outcomes at each stage. Potential failures at key steps should also be taken into
account, as they realistically represent risks and possible need for intervention.
It is also important to understand that typically CAR T-cell therapy is only recommended after
the patient has undergone previous unsuccessful treatments. The following diagram outlines the
process flow specifically related to Myeloma. It encompasses different stages, starting with the
diagnostic workup and progressing through various treatment pathways. The diagram details
the clinical decision-making process and the potential referral for CAR T-cell therapy based on
the patient's specific disease progression and response to other treatments, based on the
National Comprehensive Cancer Network guidelines for Multiple Myeloma, Version 4.2024[2].
Modeling clinical processes holds the potential for automation in practical applications.
Automating parts of the process could streamline the coordination of treatment phases, improve
the efficiency of patient data management, and ensure timely interventions. This would be
especially beneficial in complex cases where multiple treatments have been tried, and quick
decision-making is crucial. Integrating these models into healthcare systems could enhance
adherence to treatment protocols, reduced errors, and optimized patient outcomes, aligning
with the best practices set forth by the National Comprehensive Cancer Network guidelines.
Figure 2: Treatment diagram for Myeloma based on the NCCN guidelines.
2. Modeling approaches and integration of models
2.1. Methodology
In this section, the process and importance of the models is described. The process of creating
an ontological model involved defining the scope (determining the boundaries and focus of the
domain being modeled. As outlined in the article “Semantics, Ontology and Explanation” by
Giancarlo Guizzardi and Nicola Guarino[3], an important part is a process they call "ontological
unpacking," which focuses on revealing the ontological commitments of conceptual models to
enhance understanding and interoperability. The article supports defining the domain of interest
through its emphasis on identifying the relevant domain for which the ontology is to be
developed, particularly in ensuring that the ontology supports semantic interoperability tasks);
identifying key concepts and entities (enumerating the fundamental elements within the
domain) and establishing relationships (linking concepts and entities to show how they interact
or are related).
The domain of interest of the project is focused on CAR T-cell therapy. This sets the primary
domain of the ontological model. The boundaries of the domain are medical aspects (types of
cancers and subtypes), the treatments (classical and advanced treatment), and the outcomes and
side effects of the treatment.
Using OntoUML, as mentioned in the article, allows for the representation of the ontology in
a formal, machine-readable format. OntoUML is designed to provide semantically rich modeling
primitives that are aligned with ontological distinctions, which supports the creation of more
precise and expressively powerful ontological models. OntoUML is a specialized modeling
language aimed at capturing and representing ontological distinctions in conceptual models. It
draws heavily from the Unified Foundational Ontology (UFO), which provides a rich set of
ontological categories and relations. OntoUML distinguishes among various class stereotypes to
represent different types of universals that exist in reality. These stereotypes are grouped into
three main categories: Sortals, NonSortals, and Aspects[4]. The conceptual model in this article
concerns the following sortals:
Sortals are foundational for categorizing things that have a clear identity criterion, meaning
they can be individuated and counted. They include:
Kind: The most general type of thing in a domain of interest. Kinds provide the principle of
identity for their instances. An example from the ontological model is Person.
Subkind: A specialization of a kind, where the instances still comply with the identity principle
of the kind but differ in some specific characteristics. For example, Multiple Myeloma is a subkind
of the kind Myeloma.
Phase: Represents stages or temporal segments of an individual's existence that are mutually
exclusive and exhaustive for a particular kind. For example, Healthy person and Unhealthy
person could be phases of the kind Person.
Role: Temporary roles that instances of kinds or subkinds can assume during certain relations
or situations, without altering their identity. For example, Oncological patient as a role people
(Person) can assume.
Relator: Represents relational entities that mediate associations between two or more
individuals, providing the basis for their connection. An example is the Diagnosis that links a
Oncological patient to a Treatment.
The subsequent model is the PURO model. PURO is introduced as a graphical ontology sketching
approach that utilizes a first-order axiomatization. It employs a set of primitives akin to those
used in the Web Ontology Language (OWL), but with greater flexibility. In PURO, elements such
as objects, types, relations, and attributes are treated as foundational ontological distinctions
rather than merely as data modeling options. This helps in achieving a higher semantic quality
of models. Furthermore, PURO is used for creating initial sketches that can later be developed
into more rigorous reference ontologies in languages like OntoUML or into semantic web
vocabularies in OWL. Apart from its use in formal ontology engineering, PURO also serves as a
standalone graphical language for mapping out real-world situations, which is particularly useful
in discussions and explanations among human users [5]. A web-based tool called PURO Modeler
supports the PURO methodology, providing functionalities to transform sketches into different
ontology formats or to lift them to more complex models.
The final models are the MMABP models. According to the Philosophical Framework for
Business System Modeling [6] the development of an information system should be grounded in
real-world facts that exist independently, thus ensuring that the system effectively mirrors the
intricacies and operations of the business it is designed to support. The framework proposes a
four-dimensional model of the business system, which includes real world modality, real world
causality, model of collaboration and model of acting.
The real world modality represents the static view of being, detailing the system of real-world
objects and their potential relationships (the conceptual model), the model of Collaboration
captures the static view of behavior, illustrating the system of business processes and their
relationships (the process map). According to Repa, the real world causality focuses on the
temporal view of being, showing possible states in the life cycle of specific real-world objects and
the transitions between the (object life cycle). MMABP models help in optimizing and
standardizing business processes for efficiency and effectiveness. According to the article a
process map is a high-level representation of the interactions and relationships between
different business processes within an organization. It provides a global overview of the entire
system of processes, highlighting how they collaborate to achieve defined business goals.
2.2. Ontological model
The conceptual model reflects some of the principles discussed in the article "Semantics,
Ontology, and Explanation" by Giancarlo Guizzardi and Nicola Guarino. The model uses a variety
of ontological distinctions such as kinds, phases, roles, and relators, which are consistent with
the ontological theory of relations as outlined in the paper. The main concepts identified in the
CAR T-cell therapy treatment domain were cancer types, treatments, mechanisms of action,
patient characteristics and treatment outcomes.
Figure 3: Ontological model made using Menthor Editor
Using OntoUML stereotypes, these concepts have been classified:
CancerType as kind: Represents general categories of cancer that provide the identity
principle for their instances.
Treatment as kind: Diverse treatment methods form distinct categories with a clear identity.
ClassicalTreatment and AdvancedTherapy as subkind: These are specific types of treatment
methods, inheriting the identity criteria from TreatmentMethod but differing in certain
characteristics.
In the context of a CAR T-cell therapy ontological model, categorizing Treatment into
subkinds ClassicTreatment and AdvancedTreatment allows for a distinction between more
traditional forms of cancer treatment (such as chemotherapy, radiation, and surgery) and newer,
more innovative therapies like CAR T-cell therapy. CAR T-cell therapy, given its novel approach
to leveraging the body's immune system to fight cancer, falls under AdvancedTreatment. This
differentiation helps in understanding and organizing the range of treatment options available
for different cancer types within the model.
MechanismOfAction as kind: Different mechanisms by which treatments work represent
distinct categories.
TreatmentOutcome as kind: Outcomes of treatments form distinct categories with clear
identity principles.
In this model focused on CAR T-cell therapy, possible subkinds of TreatmentOutcome include
Remission, Stable Disease, Progression, and SideEffectsManaged. Remission indicates that the
cancer signs and symptoms are reduced or absent. Stable Disease means the cancer has not
significantly changed. Progression indicates that the cancer has grown or spread. Side Effects
Managed implies that any adverse reactions to the treatment have been successfully controlled.
These outcomes can help in evaluating the effectiveness of CAR T-cell therapy for different
cancer types.
TreatmentSideEffect is modeled as a separate kind with a mediation relationship to
Treatment. This distinction is important because treatment outcomes and side effects are
conceptually different aspects of the treatment process. Outcomes relate to the effectiveness of
the treatment in addressing the disease, while side effects are unintended consequences of that
treatment. By modeling TreatmentSideEffect as a separate kind with a mediation relationship to
Treatment, the model can more accurately reflect the complexity of treatment processes,
including both the intended effects (outcomes) and unintended adverse effects (side effects).
Regarding side effects, they are critical for understanding the full impact of treatment. The
approach chosen includes a kind TreatmentSideEffect with subkinds representing specific side
effects CytokineReleaseSyndrome, Neurotoxicity, and InfectionRisk. This allows the model to
encompass the range of potential adverse reactions to CAR T-cell therapy, providing a
comprehensive view of treatment implications.
Person is a kind with phases HealthyPerson and UnhealthyPerson indicating changes in state
that do not alter the underlying kind. Patient is a role that a person can take on under certain
conditions (being unhealthy). The element PatientCharacteristic has been modeled as a
perceivable quality as it represents aspects of a patient that can be observed or measured, such
as age, weight, or the presence of certain symptoms. Perceivable qualities are properties that can
be attributed to an individual and can vary from one individual to another. Modeling patient
characteristics in this way allows for a nuanced representation of the attributes that may be
relevant to their treatment or diagnosis, consistent with the principles of ontological modeling.
Diagnosis is a relator that establishes specific relations between entities like a Patient, a
CancerType, and Treatment.
Next, the relationships between these concepts were established using OntoUML's relation
stereotypes: mediation connects CancerType to Diagnosis. This is a mediation relationship
because the diagnosis mediates the effect on the cancer type, mediation also connects Treatment
to MechanismOfAction, indicating the method's mechanism.
A mediation relation is used to connect two entities that are existentially dependent on a third
entity, which is often called a relator. This relator, in the healthcare context, could be a diagnosis,
a medical construct that substantiates the relationship between a patient's condition and the
method of treatment prescribed. The diagnosis is what mediates the relationship between an
oncological patient and the treatment. The treatment is chosen based on the diagnosis, and the
appropriateness or validity of the treatment method is grounded in the diagnosis. Without the
diagnosis, the connection between the patient and the treatment method does not exist in a
justified or medically sound way.
Association links the Diagnosis to the Treatment, and also links Treatment to
TreatmentOutcome, showing the possible results of the treatment.
The specifics of the Mechanism of Action for CAR T-cell therapy can vary among different
types of cancers such as Diffuse Large B-cell Lymphoma (DLBCL), Follicular Lymphoma, Mantle
Cell Lymphoma, Multiple Myeloma, and B-cell Acute Lymphoblastic Leukemia due to the unique
characteristics of the cancer cells in each disease. For example, the antigen target for the CAR T-
cell therapy may be different, reflecting the distinct cell surface proteins found on cancer cells of
each type. The model implies that while the overarching mechanism involves immune system
activation, the precise pathway or target could differ based on the specific nature of the cancer
being treated, affecting the design and application of the therapy for each cancer type.
This approach leverages OntoUML's capabilities to model the cancer treatment domain,
providing clarity on the types of entities involved, their properties, and the complex
relationships between them.
2.3. PURO model
The PURO model represents various entities and their relationships within a CAR T-cell therapy
cancer treatment context. It includes different cancer types (Myeloma, Leukemia and
Lymphoma), as well as an instance of Follicular Lymphoma (Pat1FL). CAR T-cell therapy (a
subtype of Advanced Treatment) is connected to the cancer type subtypes through "treats"
relationships, indicating these cancers can be treated by this advanced treatment. The type
Patient has an instance Patient 1. Patient characteristics include previous treatment count and
age, as these are key attributes that impact treatment choices. Overall, the model attempts to
reflect the complexity and interrelated nature of patient characteristics, cancer types,
treatments, status in cancer care. Contrary to the conceptual model, this model does not have
TreatmentOutcome but rather DiseaseStatus. This model is only the starting point and it is
planned to build upon this model in the future, as well as alter the conceptual model to reflect
certain aspects that are and will be uncovered using PURO.
Figure 4: PURO model made using PURO Modeler.
2.4. MMABP model
The Process map offers a system-level perspective, showing the constituent processes and their
relationships and emphasizes the interrelationships between processes, showing how they work
together. Process maps help distinguish between key and support processes based on their roles
in the business system. These models are essential for understanding the broader context and
collaboration of processes within a business system, ensuring that all processes align with the
organization's goals and function cohesively.
Creating a process map for CAR T-cell therapy treatment involves visualizing the high-level
workflow of the entire treatment process, from the start event of the relapse and the patient
referral to the actual CAR T-cell therapy process. The objective of this process map is to outline
the end-to -end CAR T-cell therapy process. The models were created using TeamAssistant
software after reviewing extensive literature on CAR T-cell therapy [7][8][9][10][11][12] and
building upon the research of my master thesis[13].
The process begins with the patient’s relapse or referral from the patient’s oncologist (while
it could be argued that these events would follow one after the other, it could also be argued that
the patient could have a relapse and contact the clinic without having the referral). The CAR T-
cell therapy process involves several supporting processes (Medical History Evaluation,
Laboratory Tests, Chemotherapy, Pre-infusion Monitoring, the Infusion procedure and
Monitoring during and after infusion). After the starting event is the Patient registration,
following which is the eligibility screening that is done via the supporting processes Medical
History Evaluation and Laboratory Tests, once these are completed the Patient Screening
Assessment determines whether the patient is eligible or not eligible for treatment. If eligible for
treatment, the clinic must obtain the patient’s consent for treatment, if obtained the clinic can
schedule an appointment for the treatment. After the appointment is scheduled, there are several
possible progressions: the favorable one (that the patient shows up for the treatment, and thus
the process leads to leukapheresis), or the less favorable version (that the appointment has to
be rescheduled either due to the patient or due to the clinic) to the least favorable option (the
patient not showing up for the planned appointment without rescheduling, leading to the final
state Cancellation).
Leukapheresis can lead to three possible routes: leukapheresis being completed, the need to
redo it (or reschedule due some circumstances that however allow the possibility of
rescheduling) or a failure requiring an alternative treatment strategy. The next steps
(manufacturing, cell proliferation and quality check) all have similar three possibilities: being
unsuccessful and thus needing to repeat the leukapheresis (and scheduling the appointment to
do so), being unsuccessful but with failure requiring an alternative treatment strategy or being
successfully completed leading to the next step.
The successful quality check leads to Patient Check, this can lead to the patient having to
undergo chemotherapy (which can also lead to the patient not being able to continue treatment)
and being evaluated if still suitable further CAR T-cell therapy treatment. If the patient is suitable,
the patient will be monitored pre-infusion, the patient can be ready or not ready for infusion
(leading to treatment cancellation). If the patient is ready for infusion, what follows is the
infusion process and simultaneously patient monitoring during and directly after the infusion.
The infusion can have these possible outcomes: the patient has an acute reaction during the
infusion leading to the treatment being stopped; the infusion being unsuccessful leading to
treatment cancellation; the infusion and monitoring being successful leading to post treatment
monitoring that can have three endings: a positive response to treatment, the patient having
adverse effects (Cytokine release syndrome, neurotoxicity or other complications) or the patient
having a relapse.
2.5. Analysis and results
The paper focuses on three types of models: the ontological model provides a high-level
understanding of CAR T-cell therapy by defining key concepts and their relationships. The PURO
model provides a different look at specific elements in CAR T-cell therapy, demonstrating how
the patient, treatment methods, and outcomes interrelate. The process map offers a detailed
view of the actual treatment process, breaking down the stages into more granular steps.
The ontological model helps clarify the kinds of categories and their ties that are assumed to
exist for CAR T-cell therapy, the PURO model can serve as a tool for further ontological analysis.
During the creation process of this model, it was discussed that a more detailed PURO model
should be created in the future and also that a revised conceptual model should also be created,
reflecting the truths unearthed through the PURO model. The process map specified each step of
the CAR T-cell therapy, detailing possible outcomes (even those that are not typically mentioned
in literature concerning CAR T-cell therapy).
The models give comprehensive insights into the therapy, illustrating the end-to-end process
from the relapse to post-treatment monitoring. The models assist in understanding the therapy's
workflow, identifying challenges as well as various outcomes, and ensuring optimal treatment
strategies and answer the set questions of what are the key concepts and relationships within
the domain of CAR T-cell therapy and what are the detailed stages and interactions within the
CAR T-cell therapy workflow. The MMABP model improves the understanding and optimization
of these processes by specifying each step along with assumed possible outcomes.
Figure 5: CAR T-cell therapy process map (left) and process (right) created using TeamAssistant.
3. Discussion and conclusion
3.1. Discussion
The different models analyzed in this article contribute to understanding the CAR T-cell therapy
treatment process. The Ontological Model, PURO Model, and MMABP models each contribute to
a comprehensive understanding of CAR T-cell therapy by providing distinct perspectives and
levels of detail. While the ontological model provides a high-level conceptual framework that
identifies the primary entities and relationships within CAR T-cell therapy and serves as a
foundational framework by defining key concepts like patient, treatment types, and treatment
outcomes, it also lays the groundwork for the more detailed models, ensuring a consistent
conceptual base for all subsequent analyses. The PURO model offers a different view of the
entities, relationships, and specific elements of CAR T-cell therapy. The Process Map provides a
practical, visual guide to the workflow of CAR T-cell therapy. It translates the conceptual insights
from the Ontological and PURO Models and the simple flow chart into actionable steps. It allows
healthcare professionals to visualize the entire therapy process, identifying handoffs,
bottlenecks, and areas for improvement based on the framework provided by the other models.
The Ontological and PURO Models establish a clear conceptual and detailed framework, while
the Process Map offers a visual translation of these frameworks into the actual workflow.
Together, these models help understand the CAR T-cell therapy process from broad concepts to
specific details, offering a holistic perspective for planning, analysis, and optimization. Using
these models in conjunction provides a layered understanding of CAR T-cell therapy that can
help refine the treatment process, enhance patient care, and improve outcomes.
3.2. Conclusion
Using multiple models to map out CAR T-cell therapy aids in efficiently managing the treatment
process by clarifying the relationships between processes, understanding the therapy's
mechanisms, and ensuring patient safety. These models form a basis for creating decision-
making tools that can improve clinical outcomes.
The paper thus emphasizes the value of detailed process mapping and conceptual modeling
to improve CAR T-cell therapy's management and decision-making. Using multiple models to
map out CAR T-cell therapy aids in efficiently managing the treatment process by clarifying the
relationships between processes, understanding the therapy's mechanisms, and ensuring
patient safety. These models form a robust foundation for creating decision-making tools that
can significantly improve clinical outcomes. The comprehensive nature of this approach—
integrating ontological models, PURO models, and process maps—provides a multifaceted
perspective that is essential for enhancing precision in treatment planning and execution.
This paper underscores the critical value of detailed process mapping and conceptual
modeling to improve the management of and decision-making in CAR T-cell therapy. The insights
derived from these models not only support current clinical needs but also pave the way for the
development of advanced analytical tools that can predict treatment outcomes, customize
patient care plans, and mitigate potential risks associated with therapy. The research questions
addressed in this work aimed to clarify how fundamental concepts and relationships can be
structured and understood within the context of CAR T-cell therapy, how the detailed modelling
of the business process can provide clarity and enhance the workflow of the treatment and how
the entire modelling approach aims to identity challenges, possible outcomes and improvements
in patient management and decision-making process. The overall objective of integrating
different modelling approaches provides a more detailed understanding of CAR T-cell therapy.
Moreover, the methodologies discussed herein have broader implications for other complex
medical treatments and can be adapted to enhance systems in various therapeutic areas. Future
research should focus on refining these models through real-world data integration and
exploring their applications in other contexts to validate their effectiveness and adaptability. It
is imperative that the medical and research communities continue to collaborate in evolving
these models, ensuring they remain relevant and responsive to emerging clinical challenges.
Acknowledgments. I would like to thank professors Václav Řepa and Vojtěch Svátek for
guidance and discussions regarding the MMABP and PURO models.
Disclosure of Interests. The author has no competing interests to declare that are relevant to
the content of this article.
References
[1] National Comprehensive Cancer Network, Immunotherapy Side Effects CAR T-cell Therapy,
(2024). Accessed: April 13, 2024. [online] URL:
https://www.nccn.org/patients/guidelines/content/PDF/immunotherapy-se-car-tcell-
patient.pdf.
[2] National Comprehensive Cancer Network. Clinical Practice Guidelines in Oncology: Multiple
Myeloma. Version 4.2024, (2024). Accessed: April 13, 2024. [online] URL:
https://www.nccn.org/professionals/physician_gls/pdf/myeloma.pdf.
[3] G. Guizzardi, N. Guarino, Semantics, Ontology and Explanation, (2023).
doi:10.2139/ssrn.4673400.
[4] M. Suchanek, OntoUML Specification, (2018). Accessed: February 21, 2024. [online] URL:
https://ontouml.readthedocs.io/en/latest/classes/index.html.
[5] V. Svatek, M. Vacura, PURO: Stretching a Rope between Ontology Sketching, Foundational
Distinctions and Linked Data Vocabularies. (2024).
[6] V. Repa, Philosophical Framework for Business System Modeling. In: 2023 IEEE 25th
Conference on Business Informatics (CBI). pp. 1–6. IEEE, Prague, Czech Republic (2023).
doi:10.1109/CBI58679.2023.10187427.
[7] American Cancer Society, CAR T-Cell Therapy and Its Side Effects. Accessed: April 20 2023.
[online]. URL: https://www.cancer.org/treatment/treatments-and-side-effects/treatment-
types/immunotherapy/car-t-cell1.html.
[8] W. Aiman, M. A. Ali, H. Zubair, S. K. Butt, M.Y. Anwar, M. Hamza, N. Nanjundaiah, V. S.
Chemarthi, H. Shaaban, G. Guron, M. Maroules, F. Anwer, Advances in CAR-T Cell Therapy in
Re-lapsed/Refractory Multiple Myeloma: A Systematic Review and Meta-Analysis of Clinical
Trials in the Last Two Years, Clinical Lymphoma Myeloma & Leukemia, (2023) Vol. 23,
S485–S486. doi:10.1016/S2152-2650(23)01427-1.
[9] M. Hong, J. D. Clubb, Y. Y. Chen, Engineering CAR-T Cells for Next-Generation Cancer
Therapy. Cancer Cell, (2020), Vol. 38, no. 4, 473–488. doi:10.1016/j.ccell.2020.07.005.
[10] M. Jogalekar, R. L. Rajendran, F. Khan, C. Dmello, P. Gangadaran, B. Ahn, CAR T-Cell-Based
gene therapy for cancers: new perspectives, challenges, and clinical developments.
Frontiers in Immunology, (2022). Vol. 13, 925985. doi:10.3389/fimmu.2022.925985.
[11] M. Omar, P. Strati, L. Feng, R. Sun, A. Ayers, R. E. Steiner, R. Nair, C. Flowers, J. L. Ramdial, N.
Saini, S. A. Srour, R. E. Champlin, P. Kebriaei, L. J. Nastoupil, M. A. Rodriguez, E. J. Shpall, Y.
Nieto, J. Westin, S. S. Neelapu, S. Mandayam, S. Ahmed, Real-world analysis of safety and
efficacy of CAR T-cell therapy in lymphoma patients with decreased kidney function, British
Journal of Haematology (2023). doi:10.1111/bjh.19152.
[12] S. S. Raikar, L. C. Fleischer, R. Moot, A. Fedanov, N. Y. Paik, K. A. Knight, C. B. Doering, T.
Spencer, Development of chimeric antigen receptors targeting T-cell malignancies using two
structurally different anti-CD5 antigen binding domains in NK and CRISPR-edited T cell
lines, Oncoimmunology. (2018) Vol. 7, no. 3, e1407898.
doi:10.1080/2162402X.2017.1407898.
[13] V. Kostrouchová, The Use of Technology Intelligence Processes in the Management of
Research and Production of New Drugs, Diplomová práce. Vysoká škola ekonomická v Praze.
Vedoucí práce Jan Černý, (2023).