Cancer as a system dysfunction

Khalid Saeed, Elizabeth F. Ryder, Amity L. Manning1

Abstract

In this paper, we describe a system dynamics model that views cancer as a dysfunction of the
cellular system analogous to those of human societies. Our experiments with the model
replicate the propagation of the ailment and the impacts of the treatments. It also represents
work in progress and should be viewed as a proposition that can be further developed to
understand cancer and used to create appropriate combinations of interventions for specific
situations.

Introduction

In addition to being a condition of cells in human body, the term cancer has been used also to
describe social ailments. Corruption, crime, terrorism, have all been called cancer in the context
of how they affect other populations in society. There exist models of the interaction between
normal and rogue social institutions (e. g., Saeed and Pavlov 2008), sick and healthy populations
in the spread of infectious diseases (e.g., Thompson and Tebbens 2008) and the propagation of
chronic ailments from the perspective of patient management (e.g., Hirsch et. al. 2012), but this
perspective has not been extended to the behavior of cells in human body for disease
management. Richmond (1977) probably was the first to construct a system dynamics model to
understand the propagation of cancer, but he focused on cell deformation process rather than the
interaction of cell populations, which we attempt in this paper. Our model views cancer cells as a
part of an interconnected system of the various cell populations in an effort to understand the
propagation of the disease and the impact of the various current and experimental treatment
strategies.

1Respectively, Professor of Economics and System Dynamics, Associate Professor of Biology

and Assistant Professor of Biology, all at Worcester Polytechnic Institute. Corresponding author:
Khalid Saeed, saeed@wpi.edu

Page 1 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
Methodology

The idea of this paper originated from the first author’s brush with the ailment and the resulting
conversations with his physicians and biologist colleagues. The diverse backgrounds of the
authors subsuming environmental biology, cancer biology, economics and system dynamics,
together with the participant observer experience of one of the authors, helped to conceptualize
the structure underlying cellular interaction that was consistent both from the operational and
theoretical perspectives. Thus, the decision rules incorporated in our model are based both on
clinical and experimental explanations of cancer propagation. 2 The information base subsumes
both operational and theoretical perspectives considered important by Forrester (1994) in
creating a system dynamics model of a problem.

We view cancer as a dysfunction of the cell interactions and regulatory processes, analogous to
those of human societies that has been modeled earlier in Saeed (1986), Saeed (1990), Saeed
(1998), Saeed and Pavlov (2008) and Saeed et al. (2014). The last two papers reduce the model
of dysfunctional societies to a metaphor addressed to a class of problems rather than to a single
manifestation and serves as a template for the formulation of the model of this paper. Our model
is programmed in Stella Architect3 using system dynamics modeling framework discussed in
detail in Sterman (2000) and briefly outlined below.

Originally introduced by Jay Forrester in the1950s to address the problems of industrial

management (Forrester 1961), system dynamics came in limelight with the publication of the
controversial Limits to Growth study in 1972 (Meadows, et. al. 1972). Although this association
extended to some degree the ambivalence experienced by the Limits study also to system
dynamics, the methodology is in fact quite neutral and invaluable for exploring the behavior of a
given set of structural assumptions (Forrester 1968, Crookes and De Wit, 2014). It also allows us

2
Conversations with Drs John Phillips, Atish Chaudhary and Martin King at Dana Farber, and with Dr. Saleem
Khanani at Umas Med School were especially helpful. Additionally, Drs Amna Zarar and Erika Brutseart provided
valuable inputs that helped the initial formulation of the model.
3
A trademark of isee systems, Inc.

Page 2 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
to subsume both quantifiable and non-quantifiable variables into formal models representing a
theory with relative ease.

The underlying computational process in a model representing the microstructure of a theory can
be expressed as a set of nonlinear integral equations. System dynamics modeling allows us to
construct such models with icons and connections that can be easily assembled using dedicated
software like Stella, ithink and Vensim. 4 Table 1 shows the icons and the processes they
represent in a typical system dynamics model.

Table 1 Icons used for representing model relationships

Process Icon Explanation

Stock Accumulation or integration of
Stock 1
flows linked to the icon

Flow A rate of change or a derivative

of a stock. Empty arrowhead
indicates normal direction of
Flow 1
flow. Normally connected to a
stock. Cloud at one end
represents unlimited source or
sink
Converter Algebraic function of stocks,
other converters and constants
Converter 1

Graphical function Graphically represented

function of another variable in
graph the system

Causal link Information relationship

between two variables

4
Stella and ithink are trademark of isee systems, Inc.; Vensim is a trademark of Ventana Systems, Inc.

Page 3 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
A rectangle represents a stock that integrates the flows connected to it. A flow is a rate of change
associated with a stock, which may have more than one flows connected to it. These two types of
variables are the basic components of an abstract system implicit in a theory. Information links
from stocks to flows define decision rules. Intermediate computations transforming information
in stocks into decision rules are represented by the converter symbol. A converter is an algebraic
function of stocks, other converters and constant parameters. When an intermediate computation
involves a nonlinear graphical relationship between two variables, a tilde is added to the
converter symbol.

Once a model has been constructed using these icons, the software allows the modeler to specify
initial values of stocks, constant parameter, algebraic functions and graphical relationships that
define each flow and converter. Non-quantifiable variables are often represented by indices that
are normalized with respect to a given ambient condition. The software also allows us to
generate the behavior of these models through numerical simulation on a computer. The
simulation experiments are often designed to understand the behavioral patterns arising out of
the model structure rather than point prediction of the future (Saeed 2017).

A dynamic model of the cellular society

Our model represents the interaction between the various types of cells in an organ or a
subsystem of the body. It depicts the common structure in the simplest possible terms. It defines
a generic system that can be applied to solid as well as blood and bone-related diseases,
although we mainly address it to proliferation of sold cancers and their response to treatments in
this paper. Whether viewed in the context of an afflicted organ or a body subsystem, the
interacting cell populations can be placed in four broad categories: normal cells, pre-cancer cells,
cancer cells and immune system cells. Thus, our model contains four stocks connected by flows
shown in Figure 1.

Page 4 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
Figure 2 Key cell population stocks, their connecting flows and rules of conduct in the
cellular society.

Normal cells and Immune system cells have disciplined growth regimes in that they tend to grow
to their indicated levels – the former goal is determined by the internal intelligence of the cell,
while the later by the surveillance need created by the very existence of the unwanted cells in the
body. The unwanted cells include only pre-cancer and cancer and in our model, although this is a
simplification since other infectious and bacterial elements may also activate the immune
system. Unregulated proliferation is a hallmark of cancer (Hanahan and Weinberg, 2000) and
unlike the normal and the immune system cells where cell proliferation and turnover rates
balance to maintain their populations at indicated levels, in our model cancer cells grow
exponentially to outpace cell death.

Page 5 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
Our baseline model can be described by the following four integral equations:

normal cells = (normal cell population change - normal cell going precancerous - normal
cells killed by autoimmune disease) * dt (1)

normal cell population change = normal cell base gr + (indicated normal cells - normal
cells)/normal cell adj time (1.1)

normal cells going precancerous = normal cells * fraction normal cells going precancerous
(1.2)
normal cells killed by autoimmune disease = normal cells * fraction normal cells consumed
due to killed by autoimmune disease * immune sys capacity adequacy
(1.3)

indicated normal cells = desired normal cells (1.4)

precancer cells = (normal cell going precancerous + immune system cell transformation -
precancer cells removed - precancerous cells going cancerous) * dt (2)

immune system cell transformation = immune system cells * fraction immune cells going
precancer (2.1)

precancer cells removed = precancer cells * fr precancer cells removed * immune sys
capacity adequacy (2.2)

precancerous cells going cancerous = precancer cells * fr precancerous cells going

cancerous
(2.3)
cancer cells = (cancer cell prolifertion + precancerous cells going cancerous – cancer cells
killed and injured) dt
(3)

cancer cell proliferation = cancer cells * cancer growth fraction

(3.1)

cancer cells killed and injured = cancer cells * cancer fraction removed by immune system *
immune sys capacity adequacy (3.2)

immune system cells = (immune system cell population change - immune system cell
transformation) * dt (4)

Page 6 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
 immune system cell population change = (indicated immune system cells - immune system
cells)/immune system cell adjustment time (4.1)

indicated immune system cells = desired immune system cells = (precancer cells + cancer
cells) * immune system cell ratio (4.2)

immune sys capacity adequacy = (immune system cells / (precancer cells + cancer cells))
(4.3)

The normal and immune system cell populations remain in a dynamic balance through cellular
births and deaths, but those two rates are aggregated in the model into a single population change
rate in both cases (equations 1.1 and 4.1).

Decades of clinical and experimental evidence support a role for the immune system in
controlling cancer and have spurred clinical trials into cancer immunotherapeutic approaches.
However, recent data demonstrating that the immune system can both limit (Marin-Acevedo JA
et al., 2018) and promote (Schreiber et al 2011; Bakhoum et al., 2018) tumor growth, highlight
the complexity of the relationship between the immune system and cancer cells. Our model
recapitulates these dynamics, in part, by enabling a fraction of the pre-cancer cell population to
be killed by the immune system. The pre-cancer cell population is therefore a feature both of
transformation of normal cells to pre-cancerous cells and the level of immune system activity
(equation 2.2). When immune activity surpasses a given threshold, cancer cell growth is
enhanced instead of limited, but normal cells become susceptible to immune system surveillance,
as is seen in hyperactive immune system disorders (equation 1.3). Our model also allows some
of the immune system cells to be drawn into the precancer category as in case of normal cells,
which can happen especially in the case of aggressive cancers of blood and bone (equation 2.1),
but this rate is kept to zero in the simulations of this article, which are focused mainly on solid
cancers.

The cancer cell stock is initially populated by a transformation of precancerous cells meant to
reflect the accumulation of tumor-promoting mutations that increase with age (White et al 2014)
(equation 2.3). Once transformed, the cancerous cells exhibit unregulated proliferation and thus
cancer grows exponentially (equation 3.1). The immune response increases to combat the rising

Page 7 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
cancer cell population (equation 4.2), and it is ultimately the imbalance of tumor growth vs
immune response that yields a proliferation of cancerous cells.

These processes, incorporated into our model, are widely recognized but are addressed in
conceptual silos rather as an interconnected system. Medical interventions based on a stand-
alone view of a process can lead to unintended consequences widely experienced in most
treatment regimes and tolerated as unfortunate but unavoidable side effects. We have attempted
to integrate these processes into an interconnected system to be able to better understand the
dynamics of cancer cell growth and the impacts of its treatments and their side effects.

Model behavior

The model is supplied with an internally consistent set of parameters and initial conditions so
under normal conditions, all cell populations except cancer are maintained in a dynamic
equilibrium or homeostasis. Cancer population is initialized at zero value. The simulation time is
set at 1000 months (83.3 years) that approximates the average life expectancy of a healthy
individual. It should be noted that despite the equilibrium, all cell populations constantly turn
over rather than remaining constant. The in- and out- flows tied to each stock continue while the
stocks remain in a dynamic balance.

When initially populated by the transformation of pre-cancer cells, the cancer cells begin to grow
exponentially. Low and intermediate rates of growth of cancer are contained however by a
concomitant increase in the immune activity that constantly kills the cancer cells and keeps their
population under control. The absolute number of cancer cells is influenced by cancer cell
proliferation, as well as the rate at which normal cells acquire pre-cancerous characteristics and
then are ultimately transformed into cancer cells. To model rates of proliferation that may differ
between aggressively growing and slow growing tumors, the model allows for modulation of
cancer cell proliferation rates. To allow incorporation of the potential impact of carcinogenic
exposure, the model also allows for modulation of the rate at which normal cells are transformed
to cancer cells. Simulations showing system response to the various rates of increase in cancer

Page 8 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
cell number are shown in Figure 3. As expected, the model response shows that low growth and
transformation rates can be contained by the immune system, while cancer cells are able to grow
exponentially when immune cell growth and/or transformation exceeds the expansion in the
immune system capacity. At this time, we have not attempted to precisely calibrate the model
and our experiments attempt only to qualitatively verify the sensitivity of the cell system to the
parameter representing cancer growth rate.

cancer cells
100
6

50 7 5

5
6

5
4 3
0 2 3 1 2 1
1
0 250
1 8% cancer growth
4 9.5% cancer growth
7 11% cancer growth
4 4 4
3 3
2 1 2
500 750 1000
months
2 8.5% cancer growth 3 9% cancer growth
5 10% cancer growth 6 10.5% cancer growth

Figure 3 Comparative plots showing a tipping point in cancer containment

Since slow cancer growth rates can be contained by the immune system, they may not offer
fatality risk over the lifetime and need not be treated. However, as genetic, environmental, or
behavioral factors may alter this balance, cancer patients must be monitored carefully even when
the treatment option is not taken. A faster proliferation rate calls for immediate intervention.

Page 9 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
Cancer aggressiveness is often determined through a pathological examination of samples
extracted through a biopsy (Weidner 1995). Specific indicators differ for different types of
cancers. For example, Gleason scores that measure cellular differentiation, cancer cell infiltration
into normal tissue, and cellular proliferation, are used to assess the aggressiveness of prostate
cancer in a pathological examination (Gleason 1977).

Testing hypotheses about lifetime risk of cancer

Several hypotheses have been advanced about the risk of onset of cancer over an individual’s
lifetime. These include 1) a gradual increase in transformation of precancer cells into cancer; 2) a
decline in the efficacy of the immune system such that immune cells are present but less
effective at clearing pre cancer and cancer cells, 3) a slowing down of the adjustment of the
immune system cell population to a needed level and 4) a gradual reduction in the population of
immune system cells (White et al 2014, Montecino-Rodriguez et al 2013).

These hypotheses can be translated into metrics affecting our model parameters that may change
gradually over time. Figure 4 shows the growth of cancer in response to the lifetime changes
suggested by these hypotheses compared to a base case that subsumes a slow transformation of
precancer to cancer with no change in immune system response. Trend slopes are arbitrarily
chosen to accommodate all cases in the same ballpark range.

Page 10 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
 cancer cells
200

100
4
2 3

5
2 3 4 1
5 1
2 3 4
5 1
0 3 4
1 2
0 250 500 750 1000
months
1 base case lifetime cancer risk
2 lifetime increase in precancer transformation to cancer
3 lifetime decline in immune system effectiveness
4 lifetime slowing of immune system population adjustment
5 lifetime decline in immune system population goal

Figure 4 Testing hypothesis about lifetime changes in precancer transformation and

immune system response

All simulations pertaining to the four hypotheses show a rising trend that becomes uncontrollable
over time, the rate of rise depending on the slope of the lifetime trends, which can be adjusted to
reflect genetic instability as influenced by genetic and environmental factors (ie cigarette
smoking, pollution, UV and other carcinogen exposure).

Cancer treatments and their performance

Most treatments for solid cancer aim at either surgically removing the affected organ or a part of
it, subjecting it to drugs (nutrient blockers as well as lethal chemicals) and external or internal
radiation. Hematologic and Musculoskeletal cancers have been treated with chemotherapy and
most recently by immunotherapy.

Page 11 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
These treatments are often administered in combination based on the pathological stage and
genetic makeup of the cancer. Conservative surgical approaches are limited by a desire to obtain
complete resection of the tumor, while minimizing collateral damage to healthy tissue. Similarly,
chemotherapeutic approaches must balance efficacy in killing cancer cells and potential off
target effects on healthy cells. Even so, tumor relapse and drug resistance remain challenges in
the clinic. Figure 5 compares the impact of prevalent treatments with different effectiveness
targets using our model and assuming reasonable achievement of intended targets. All treatments
are invoked at month 700 and last for 12 months.

Figure 5 Simulation of treatments with different targets

Graph 1 represents the baseline simulation showing the average cancer growth risk over lifetime
due to an increasing transformation of precancer cells to cancer, all other assumptions subsumed
in equations 1-4 remaining unchanged. Graph 2 simulates a treatment regime implemented at
month 700 on top of this trend, lasting for 12 months and eliminating 99% of the malignancy
each month, which is median achievable target of the treatment even if the treatment intent is to
eliminate 100% of cancer. The sensitivity of achievement fraction can be tested. In this simple
model, it would change the initial decline. The simulation shows that cancer resumes its
exponential growth profile after the treatment ends, although it might impose a lower threat for
the patient since the resurging cancer cell population is now much smaller. The model assumes
that treatment eliminates 99% of the cancer cells, but that remaining cells can resume

Page 12 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
proliferation after treatment concludes. Two potential outcomes not modeled here include that 1)
the remaining cells are injured and disabled such that their proliferative capacity is drastically
reduced (the preferred outcome), and 2) that the remaining cells have acquired resistant to the
therapeutic intervention and as such may be resistant to subsequent therapeutic approaches
(Housman et al., 2014)

Graph 3 shows an interesting treatment scenario whereby reduction of cancer cell proliferation,
in the absence of therapy-induced cancer cell death is sufficient for complete and sustained
remission. In this scenario, the immune system is able to keep cancer cell proliferation at bay
following therapeutic intervention to reduce cancer cell proliferation by only 20%. Graph 4
models off-target effects of the therapeutic intervention in Graph 3 such that normal and immune
cell replenishment is also reduced by 20%. showing such damage has only a moderate impact on
the timing of remission. Populations of normal and immune cells recover after the treatment
concludes since both populations must adjust to their respective indicated values.

Preclinical approaches to identify novel therapeutics often focus on agents that achieve
maximum cancer cell death, with the expectation that efficacy of a therapeutic approach
correspond to the cytotoxicity of the drug. Our simulation however suggests instead that
approaches to reduce cancer cell proliferation, when coupled with an appropriate immune
response, may also be an effective treatment strategy. Thus, treatments aimed at altering
proliferation, without inducing cell death warrant further exploration.

The model also allows testing of immunotherapy regimes. Two options were tested: 1)
increasing immune cell potency, which has recently been attempted through targeted
immunotherapies (Marin-Acevedo et al., 2018) and 2) stimulating the immune system in general
(Bast 1976), one manifestation of which is the non-specific immunotherapies which include use
of drugs like Cytokines that help immune system cells grow and divide more quickly5.

Figure 6 shows sensitivity runs comparing the two approaches to immunotherapy. It is observed
that impact on cancer cell population is quite comparable in the two approaches. In both cases,

5https://www.cancer.org/treatment/treatments-and-side-effects/treatment-
types/immunotherapy/nonspecific-immunotherapies.html

Page 13 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
the treatment becomes effective after a threshold where the cancer kills affected by the immune
system response can exceed the cancer proliferation rate. Cancer growth is able to overcome a
marginal change (5%) in either the immune system potency or the population of the immune
system cells. However, a moderate change in immune system behavior (10%) can reverse the
cancer growth trend leading to sustained remission.

sensitivity of treatments rasing immune cell potency

500
millions of cancer cells

250

1
2

5
4 3
3
2
1
4 5 4
3
0 1 2 3 4 5 1 2 5

0 250 500 750 1000

months
1 base run with no treatment
2 treatment creating 5% increase in immune system cell potency at month 700
3 treatment creating 10% increase in immune system cell potency at month 700
4 treatment creating 15% increase in immune system cell potency at month 700
5 treatment creating 20% increase in immune system cell potency at month 700

sensitivity of treatments rasing immune cell population

500
millions of cancer cells

250

1
2

5
4 3
3
2
1
4 5 4
3
0 1 2 3 4 5 1 2 5

0 250 500 750 1000

months
1 base run
2 treatment creating 5% increase in immune system cell ratio at month 700
3 treatment creating 10% increase in immune system cell ratio at month 700
4 treatment creating 15% increase in immune system cell ratio at month 700
5 treatment creating 20% increase in immune system cell ratio at month 700

Figure 6 Sensitivity simulations comparing immunotherapy options

Page 14 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
A moderate to high change (20%) in immune system potency (as in targeted immunotherapy
regimes) and immune system population (as in nonspecific immunotherapy regimes) can have an
undesirable and deleterious impact on other body systems, contributing to autoimmune
responses. Figure 7 compares the level of automimmune response for the two immunotherapy
options. Consistent with the clinical data, our model indicates that such therapies may create
increased risk for autoimmune disorders. And that there is a need to explore ways to instead
enhance the immune system potency, such as cancer vaccines, immune checkpoint inhibitors and
monoclonal antibodies.

autoimmune impact of alt immunotherapy regimes

1.3
millions of normal cells under lupus attack

3
3

1.2

1.1

1 2 3 1 2
2 3 1 2
1 1

0.9
200 400 600 800 1000
months
1 baeline immune system activity
2 immune system activity with 20% incrase in immune system potency
3 immune system activity wit 20% increase in immune system cell population

Figure 7: Autoimmune disorder risk for alternative immunotherapy regimes

Model limitations

Our model so far integrates fragmented concepts from literature but has many limitations. First
of all, the various cell populations are treated as undifferentiated aggregates, whereas in reality,

Page 15 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
many types of cells behaving differently to the stimuli and the treatments exist and offer many
challenges to the treatment process. We also have not tested sequential treatment regimes that
target specific ctos-sections of cancer and are often used to combat complex conditions. We also
have not addressed the process of systematic deterioration of the body functions as cancer
proliferates. This deterioration introduces an element of path dependence in the treatment
regimes and its consideration would require understanding the role of the shared common
nourishing environment like the blood supply.

Even though the biological organisms are open systems that draw energic inputs from their
environment and can produce more nourishment when their need increases (Bertalanffy 1968),
their processing capacity that creates the shared nourishment environment does offer a limit.
Thus, the growing cancer and immune cell response will ultimately be affected by the body’s
shared nourishment capacity, a manifestation of which is the blood supply. Excessive use of the
shared nourishment by cancer will inhibit the maintenance of the human body and the failure and
corruption of the immune system in advanced stages of cancer. We are however unable to verify
our constructs on the role of nourishment environment in the propagation and control of cancer
from the available literature, which seems to posit that even though the cancer cells have a higher
metabolic demand, nutrients are not a limiting factor and so cancer cell proliferation is not
sensitive to nutrient supply (Liberti and Locasale 2016).

Given the normal cells create a nourishing environment for all cell populations that sustains their
functions, an antagonist relationship between normal cells and cancer cells in terms of sharing
the body’s nourishing commons must be taken into account. The healthy cells of the body might
be able to transform more external nutritional inputs into nourishment, they would albeit be
ultimately limited by capacity constraints as manifest in blood volume and organ and subsystem
functions. Thus, even when cancer cells are able to create their exclusive blood supply, they
would diminish the inclusive environment that nourishes this supply is shared by all cells. A
nourishment capacity limitation therefore makes intuitive sense although it may invoke different
responses from healthy and cancer cells. It may therefore be interesting to model how
angiogenesis will promote tumor growth, or how antineoangionesis therapy might limit growth.

Page 16 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
Conclusion

We have made a preliminary attempt in this paper to model the development of cancer as an
interaction between normal, immune system, pre-cancer and cancer cell populations. Various
prevalent and experimental treatment regimes are tested and their unintended consequences
explored. While reasonable parameter values are used to experiment with the model, our analysis
in this first effort remains qualitative. Treatments aimed at limiting replication ability of cancer
and targeted immunotherapies seem to be promising.

Our model could be extended to create disaggregation within each cell population category so a
larger variety of treatment regimes affecting different types of cells could be tested. Also, while
we see the role of inclusive nourishment resources of the body, such as blood supply, to be an
important part of the cell interaction in the body, it is not yet included in the model. We,
however, recognize the importance of the inclusive nourishment commons of the body embodied
in its blood supply but defers it to future work.

References

Arta M. Monjazeb AM, Hui-Hua Hsiao HH, Gail D. Sckisel GD, WJ. 2012. The role of antigen-
specific and non-specific immunotherapy in the treatment of cancer. Journal of
Immunotoxicology, 9(3): 248-258, DOI: 10.3109/1547691X.2012.685527
Bakhoum SF, Ngo B, Laughney AM, Cavallo JA, Murphy CJ, Ly P, Shah P, Sriram RK,
Watkins TBK, Taunk NK, Duran M, Pauli C, Shaw C, Chadalavada K, Rajasekhar VK,
Genovese G, Venkatesan S, Birkbak NJ, McGranahan N, Lundquist M, LaPlant Q, Healey
JH, Elemento O, Chung CH, Lee NY, Imielenski M, Nanjangud G, Pe’er D, Cleveland DW,
Powell SN, Lammerding J, Swanton C, Cantley LC. 2018. Chromosomal instability drives
metastasis through a cytosolic DNA response. Nature. 553: 467–472
Bast RC. (09/1976). Critical Review of Previously Reported Animal Studies of Tumor
Immunotherapy with Non-Specific Immunostimulants. Annals of the New York Academy of
Sciences (0077-8923), 277 (1): 60.
Bertalanffy LV (1968) General Systems Theory, New York: George Braziller.
Crookes DJ, De Wit MP. 2014. Is system dynamics modeling of relevance to neoclassical
economists? South African Journal of Economics 82(2): 181–192.
Forrester JW. 1961. Industrial Dynamics. Cambridge, MA: MIT Press.
Forrester JW. 1968. Principles of Systems. Cambridge, MA: Wright-Allen
Forrester JW. 1994. System dynamics, systems thinking, and soft OR. System Dynamics
Review.10(2-3): 245-256

Page 17 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843
Gleason, DF. 1977. The Veteran's Administration Cooperative Urologic Research Group:
histologic grading and clinical staging of prostatic carcinoma. In Tannenbaum M. Urologic
Pathology: The Prostate. Philadelphia: Lea and Febiger. pp. 171–198
Hanahan D, Weinberg RA. 2000. The hallmarks of cancer. Cell. 100(1): 57-70.
Hirsch G, Saeed K, McCleary K, Myer K. 2012. Deceased Donor Potential for Organ
Transplantation: A System Dynamics Framework. 30th annual conference of the
International System Dynamics Society. St. Gallen Switzerland: System Dynamics Society.
Housman G, Byler S, Heerboth S, Lapinska K, Longacre M, Snyder N, Sarkar S. 2014. Drug
Resistance in Cancer: An Overview . Cancers. 6(3), 1769-1792
Liberti MV, Locasale JW. 2016. The Warburg Effect: How Does it Benefit Cancer Cells? Trends
Biochem Sci. 41(3): 211–218.
Marin-Acevedo JA, Bhagirathbhai D, Soyano ES, Knutson KL, Chumsri S, Lou Y. 2018. Next
generation of immune checkpoint therapy in cancer: new developments and challenges.
Journal of Hematology & Oncology. 39(11):1-20
Meadows DH, Meadows DL, Randers J, Berhens III BW. 1972. Limits to Growth. Washington
DC: Potomac.
Montecino-Rodriguez, E., Berent-Maoz, B., & Dorshkind, K. (2013). Causes, consequences, and
reversal of immune system aging. The Journal of Clinical Investigation, 123(3), 958–965.
http://doi.org/10.1172/JCI64096
Richmond BM. 1977. Towards a Structural Theory of Cancer. MIT: System Dynamics Group
Memo D-2718
Saeed K, Pavlov O, Skorinko J, Smith A. 2014. Farmers, Bandits and Soldiers: A generic system
for addressing peace agendas. System Dynamics Review. 29(4): 237–252
Saeed, K. 1986. The Dynamics of Economic Growth and Political Instability in the Developing
Countries. System Dynamics Review. 2(1).
Saeed, K. 1990. Government Support of Economic Agendas in Developing Countries: A
Behavioral Model. World Development. 18(6).
Saeed, K. 1998. Maintaining Professional Competence in Innovation Organizations. Human
Systems Management. 17(1): 69-87.
Saeed, K. 2017. Circumscribing System Dynamics Modeling and Building Confidence in
Models: A Personal Perspective. Working paper. Available at
SSRN: https://ssrn.com/abstract=3093080
Saeed, K. and O. Pavlov. 2008. Dynastic cycle: A generic structure describing resource
allocation in political economies, markets and firms. Journal of Operations Research
Society. 59(10): 1289-1298.
Sterman JD. 2000. Business Dynamics, Systems Thinking and Modeling for a Complex World.
Boston: Irwin McGraw-Hill
Thompson KM, Tebbens RJD. (2008). Using system dynamics to develop policies that matter:
global management of poliomyelitis and beyond. System Dynamics Review 24(4): 433–449.
Weidner N. 1995. Current pathologic methods for measuring intratumoral microvessel density
within breast carcinoma and other solid tumors. Breast Cancer Research and Treatment.
36(2): 169–180.
White MC, Holman DM, Boehm JE, Peipins LA, Grossman M Henley J. 2014. Age and Cancer
Risk: A Potentially Modifiable Relationship. Am J Prev Med._46(301_):_S7–15_.

Page 18 of 18

Electroniccopy
Electronic copyavailable
available at:
at: https://ssrn.com/abstract=3235843
https://ssrn.com/abstract=3235843