Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Mathematical modelling as a proof of concept for MPNs as a human inflammation model for cancer development

  • Morten Andersen,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Department of Science and Environment, Roskilde University, Roskilde, Denmark

    ⨯
  • Zamra Sajid,

    Roles Formal analysis, Software, Writing – review & editing

    Affiliation Department of Science and Environment, Roskilde University, Roskilde, Denmark

    ⨯
  • Rasmus K. Pedersen,

    Roles Formal analysis, Software, Writing – review & editing

    Affiliation Department of Science and Environment, Roskilde University, Roskilde, Denmark

    ⨯
  • Johanne Gudmand-Hoeyer,

    Roles Formal analysis, Software, Writing – review & editing

    Affiliation Department of Science and Environment, Roskilde University, Roskilde, Denmark

    ⨯
  • Christina Ellervik,

    Roles Conceptualization, Methodology, Writing – review & editing

    Affiliation Department of Laboratory Medicine at Boston Children’s Hospital, Boston, Massachusetts, United States of America

    ⨯
  • Vibe Skov,

    Roles Conceptualization, Methodology, Validation, Writing – review & editing

    Affiliation Department of Hematology, Zealand University Hospital, University of Copenhagen, Roskilde, Denmark

    ⨯
  • Lasse Kjær,

    Roles Conceptualization, Methodology, Validation, Writing – review & editing

    Affiliation Department of Hematology, Zealand University Hospital, University of Copenhagen, Roskilde, Denmark

    ⨯
  • Niels Pallisgaard,

    Roles Conceptualization, Methodology, Writing – review & editing

    Affiliation Department of Pathology, Zealand University Hospital, University of Copenhagen, Roskilde, Denmark

    ⨯
  • Torben A. Kruse,

    Roles Conceptualization, Methodology, Writing – review & editing

    Affiliation Department of Clinical Genetics, Odense University Hospital, Odense, Denmark

    ⨯
  • Mads Thomassen,

    Roles Conceptualization, Methodology, Writing – review & editing

    Affiliation Department of Clinical Genetics, Odense University Hospital, Odense, Denmark

    ⨯
  • Jesper Troelsen,

    Roles Conceptualization, Methodology, Writing – review & editing

    Affiliation Department of Science and Environment, Roskilde University, Roskilde, Denmark

    ⨯
  • Hans Carl Hasselbalch ,

    Roles Conceptualization, Investigation, Methodology, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

    hans.hasselbalch@gmail.com

    Affiliation Department of Hematology, Zealand University Hospital, University of Copenhagen, Roskilde, Denmark

    ⨯
  • Johnny T. Ottesen

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Department of Science and Environment, Roskilde University, Roskilde, Denmark

    ⨯

Abstract

The chronic Philadelphia-negative myeloproliferative neoplasms (MPNs) are acquired stem cell neoplasms which ultimately may transform to acute myelogenous leukemia. Most recently, chronic inflammation has been described as an important factor for the development and progression of MPNs in the biological continuum from early cancer stage to the advanced myelofibrosis stage, the MPNs being described as “A Human Inflammation Model for Cancer Development“. This novel concept has been built upon clinical, experimental, genomic, immunological and not least epidemiological studies. Only a few studies have described the development of MPNs by mathematical models, and none have addressed the role of inflammation for clonal evolution and disease progression. Herein, we aim at using mathematical modelling to substantiate the concept of chronic inflammation as an important trigger and driver of MPNs.The basics of the model describe the proliferation from stem cells to mature cells including mutations of healthy stem cells to become malignant stem cells. We include a simple inflammatory coupling coping with cell death and affecting the basic model beneath. First, we describe the system without feedbacks or regulatory interactions. Next, we introduce inflammatory feedback into the system. Finally, we include other feedbacks and regulatory interactions forming the inflammatory-MPN model.

Using mathematical modeling, we add further proof to the concept that chronic inflammation may be both a trigger of clonal evolution and an important driving force for MPN disease progression. Our findings support intervention at the earliest stage of cancer development to target the malignant clone and dampen concomitant inflammation.

Citation: Andersen M, Sajid Z, Pedersen RK, Gudmand-Hoeyer J, Ellervik C, Skov V, et al. (2017) Mathematical modelling as a proof of concept for MPNs as a human inflammation model for cancer development. PLoS ONE 12(8): e0183620. https://doi.org/10.1371/journal.pone.0183620

Editor: Pal Bela Szecsi, Holbæk Hospital, DENMARK

Received: March 16, 2017; Accepted: August 8, 2017; Published: August 31, 2017

Copyright: © 2017 Andersen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: The authors received no specific funding for this work.

Competing interests: C. Ellervik, V. Skov, L. Kjær, T.A. Kruse, M. Thomassen, M. Andersen, J.T. Ottesen, J. Troelsen, Z. Sajid, R.K. Pedersen and J. Gudmand-Hoeyer have no conflicts of interest to disclose. N. Pallisgaard has the following to disclose: consulting or advisory role for Qiagen and Novartis, participated in speakers bureau (Novartis and Roche), research funding (Incyte and Novartis), holds a patent (patent pending on contamination control of lymphocyte DNA when measuring cell free DNA in liquid biopsies), travel and accommodation (Novartis). H.C. Hasselbalch has the following to disclose: invited speaker and advisory role (Gilead), research funding (Novartis), travel and accommodation (Gilead, PharmaEssentia, Novartis). This does not alter our adherence to PLOS ONE policies on sharing data and materials.

Introduction

The classic chronic Philadelphia-negative myeloproliferative neoplasms (MPNs) include essential thrombocythemia (ET), polycythemia vera (PV) and primary myelofibrosis (PMF), which are acquired stem cell neoplasms [1]. Most patients live with their MPNs for decades although with a huge morbidity burden due to a high risk of thrombosis with cardiovascular complications and a massive comorbidity burden as well due to an increased propensity to develop autoimmune and chronic inflammatory diseases [2–4], including a 40% increased risk of second cancers [5,6]–not only after the MPN-diagnosis but also prior to the MPN-diagnosis [7]. Several years prior to the MPN-diagnosis these patients also have an increased risk of cardiovascular, autoimmune and inflammatory diseases [8,9]. Furthermore, the MPNs have an inherent risk of transformation to acute myelogenous leukemia (AML) and myelodysplastic syndrome (MDS) [10].

During the last decade major breakthroughs have occurred in the understanding of the pathogenesis of the MPNs, the most important being the identification of the somatic clonal markers–JAK2, MPL and CALR [11–18]. The findings of several other mutations already at the time of MPN-diagnosis, with the emergence of additional mutations in the advanced transforming stages of MPNs [17,18], all support the concept of a biological continuum from the early cancer stages (ET/PV) to the advanced cancer stages (myelofibrosis or AML) [1,19,20]. Chronic inflammation is the common link between common diseases such as atherosclerosis, the metabolic syndrome, type II diabetes mellitus and cancer, in which the JAK-STAT- signalling and the NF-kB pathways are activated and have major roles in disease progression [21–28]. These pathways are activated in MPNs as well. Most recently, the MPNs have been described as “Inflammatory Diseases “[4] and “A Human Inflammation Model For Cancer Development”[29] reflecting chronic inflammation to be a major driving force for clonal evolution and disease progression in MPNs [30–39]. This novel concept is built upon a platform, which has combined data from studies in several research fields and disciplines within MPNs—clinical [3–9,29–53], experimental [54–63], genomic [64–70], immunological [71–74] and not least epidemiological studies [3,5–7,75–77].

Another research field—mathematical modelling of cancer development—has not been applied to a similar extent within MPNs until very recently [78,79] and not in the context of investigating the concept of MPNs as “A Human Inflammation Model for Cancer Development”. Mathematical modelling of cancer development has provided new insights regarding cancer initiation and progression [80–89]. In this context, mathematical modelling has a huge potential to support or disprove understanding of research data on pathogenetic factors of significance for cancer development but also in regard to providing supportive evidence for a drug to be used in cancer therapy and accordingly a novel tool in evidence-based medicine [90–92]. Mathematical modelling of chronic inflammation as the trigger and driver of MPNs has never been investigated. Although the concept of MPNs as “inflammatory diseases” is being increasingly recognized, additional proof of this novel concept by mathematical modelling might be of utmost importance not only for our understanding of the pathogenesis of these neoplasms, but also in regard to diagnosis and treatment. Herein, we for the first time by mathematical modelling add further proof of the concept that MPNs may be both triggered and driven by chronic inflammation. We discuss the perspectives of our findings, which might implicate intervention at the earliest stage of cancer development (ET, PV) to target the malignant clone and dampen concomitant inflammation when the tumor burden is minimal, and accordingly, the outcome of treatment is logically most favorable.

Methods

The system describes the proliferation from stem cells to mature cells including mutations of healthy stem cells to become malignant stem cells. We include regulatory interactions (e.g. niche growth effects) and inflammation coping with cell death, inflammatory cytokines, and neutrophils. In order to design an inflammatory MPN model, we build on the coupled dynamics of inflammation and cancer progression as depicted in Fig 1.

thumbnail
Download:
Fig 1. The conceptual model.

Light gray boxes (symbolized x0, x1, y0, and y1) illustrate the compartments of the basic model, and the black arrows the rates of the flows between these compartments. Here x0 denotes the number of HSC, x1 that of HMS, y0 that of MPN SC, and y1 the number of MPN MC. The light blue compartment (symbolized a) contains all dead cells and the light orange compartment (symbolized s) the inflammatory level, i.e. the immune response. Blue arrows from these represent related rates of flows. Red stipulated arrows going from the inflammatory compartment represent effects of the cytokines (or neutrophils when eliminating dead cells) modulating rates of the basic model. Two additional rates (depending on x0 and y0) appearing as red stipulated arrows represent the bone marrow niches symbiosis with the stem cells modulating the self-renewal rates. Note, stem cells leaving their respective compartments enter the corresponding mature cell-pools as multiplied by the progenitor amplification factor (A).

https://doi.org/10.1371/journal.pone.0183620.g001

The model

Most previous studies attempting to model the role of inflammation and immune deregulation in cancer progression consider solid tumors and couple the T-cell and natural killer (NK) cell dynamics to a logistic growth of a tumor. They mainly describe quite simplified versions of the adaptive immune response without explicitly considering the underlying cancer growth dynamics [93–98]. In contrast to all these models, our model is the first which couples the principles underlying actual cell dynamics to a basal inflammatory response. This response is seen for a normal infection, where the amount of dead cells provokes the immune response and stimulates the renewal of stem cells. Despite this complex coupling, the model is kept as simple as possible still allowing the relevant quantities to be described. Thus, the goal is to describe an important coupling between MPN development and the inflammatory response at a quantitatively conceptual level. Hence, the complicated mathematical question of model identifiability and accurate parameter estimation will be addressed elsewhere. Nevertheless, we include some model calibration and validation after presenting the model to justify and demonstrate the strength of the model.

Basically, our model consists of four pools of cells; the hematopoietic stem cells (HSC), the hematopoietic mature cells (HMC), the MPN-mutated stem cells (MPN SC) and the MPN mature cells (MPN MC). The number of these cells are denoted x0, x1, y0, and y1 respectively, where x refers to normal hematopoietic cells and y to MPN hematopoietic cells, while index 0 refers to stem cells and index 1 to mature cells. A single stem cell (SC) may proliferate in three ways; symmetric self-renewal (having two stem cells as offspring), asymmetric self-renewal (turning into one stem cell and one progenitor cell), and symmetric differentiation (giving rise to two progenitor cells). The progenitor cells cannot be ignored, however, we consider the progenitor cells simply as intermedia multiplication steps describing the way stem cells generate mature cells. In the model, the generations or continuum of progenitor cells will be implicitly accounted for as each stem cell will generate a number of mature cells by an amplification factor, A (= 2k if there are k generations of progenitor cells). Feedbacks from or to progenitor cells are ignored or integrated into the other included feedbacks.

The present focus is on the ensemble of each cell type and not the individual cells; thus the governing laws will be for the pools of cells, in science denoted compartments [99]. Mathematically, the dynamics will be described by non-linear ordinary differential equations respecting conservation laws. The HSC self-renews with rate rx and the malignant MPN SC self-renews with rate ry. Furthermore, HSC may be transformed by cell division by a rate ax whereas the MPN SC does so with a rate ay. The mature cells are multiply generated, i.e. the HMC are generated with a rate ax·Ax and the MPN MC with a rate ay·Ay. Finally, all cell types may die; stem cells with a lower rate and mature cells with a higher rate. The turnover (or mortality) rates are dx0, dx1, dy0, and dy1 for the HSC, HMC, MPN SC, and MPN MC, respectively. Except for the mutation part and the multiplication factor, this duplicates the structure of the model proposed by Dingli and Michor (they silently used A = 1) [92].

A small probability rm describes the mutation of HSC into MPN SC. In that case, rm is not the probability of a single mutation but possibly a serial sequence of mutations turning the HSC into a cancer cell capable of self-renewal, by definition an MPN SC, where a mutation is expected to be described by a Poisson process [100]. The probability for one mutation is about 10−7 per year per cell [101]. However, not all mutations are malignant; only mutations which happen on particular locations (i.e. at specific nucleic acids) of the DNA cause MPN relevant mutations. Inflammation increases the risk of mutations, including smoking, exposure to ultraviolet light or certain chemicals [49,50,101–104]. It is this small probability which violates a possible deterministic description with a simple mutation rate. Except for the mutation part, the model will be deterministic and continuous. In most of our work, we studied the development right after the first malignant mutation has occurred (denoted the first insult). In these cases, the simulations start with one malignant stem cell. Meanwhile, the number of all other cells are in a healthy steady state with the mutation rate put to zero. The approach is justified by the fact that including a non-zero mutation rate did not affect the outcome of the model.

The equations are all of the general form, resulting in specific systems of ordinary differential equations as given in S1 Appendix.

Whenever cells die the debris have to be engulfed by phagocytic cells, e.g. neutrophils and macrophages while a hierarchic cascade of pro- and anti-inflammatory cytokines are released [96–98,110]. Following the parsimonious principle, we let the dead cells (a) up-regulate the amount of phagocytic cells (s) with rate constant rs per dead cell while they are eliminated with a rate es. In addition, endotoxins, smoking and other environmental factors may add to the inflammatory response; thus we add such a term (characterized by the lightning symbol in Fig 1). Since MPNs develop on time-scale years and inflammatory immune processes are fast (on time-scale hours-days), we assume that the amount of phagocytic cells is balanced by the cytokines levels in a fixed ratio. Thus, the cytokine level is proportional to the phagocytic level why the inflammatory compartment (s) represents both (up to a possible proportionality constant which may be incorporated into the rate constants). Meanwhile, the amount of dead cells is down-regulated as a second order elimination process, -ea·a·s, with rate constant ea. Dead cells are produced by d0·x0+dy0·y0+dx1·x1+dy1·y1 per time denoted the turnover, which is assessed by the plasma concentration of lactic dehydrogenase (LDH). It is well-known that the inflammatory level affects the mutation rate [104] and the self-renewal rates [105]. For simplicity, we take these to be proportional with the inflammatory level (of course saturation may occur) but since the level (a) settles at constant levels so does the inflammatory level (s), which may be thought of as the amount of inflammatory cytokines which have been shown to be increased in patients with MPNs and several in a step-wise manner from controls over the early cancer stages (ET, PV) to the advanced cancer stage–myelofibrosis (PMF) (S1 Appendix) [40–46]. Thus, it turns out that various specific cytokines (IL-1β, IL-1RA, IL-2R, IL-6, IL-8, IL-10, IL-12) and C-reactive protein (CRP)—a conventional biomarker of inflammation—are linearly correlated with the inflammatory level (s). These cytokines have been chosen for validation of our model since elevated levels of several of these cytokines have been associated with an inferior survival [44]. Likewise, elevated levels of CRP have been shown to be associated with shortened leukemia-free survival in patients with myelofibrosis [42]. Of note, the inflammatory cytokine IL-8 have been reported to be of particular interest in the context of MPN pathogenesis [57–60]. These extra pools of cells are depicted in Fig 1 along with the rates governing the dynamics. This establishes the coupled inflammatory-MPN model. The full system of mathematical equations, representing the model is described in Table B in S1 Appendix including default parameter values.

Model calibration, validation, and results

The model is inspired by Dingli and Michor, and therefore the parameter values are based upon their values [92]. However, we have adjusted them to obtain more appropriate saturation levels in agreement with data (see Fig 2 and the reported values in Table C in S1 Appendix). First, the model is calibrated to the situation of no MPN cancer cells (y0i = 0 and y1i = 0). In this situation, we expect a stable steady state such that the number of HSC is approximately 104 and that of HMC is approximately 1010. These choices are compromises between reported values for the number of HSC [78, 86, 88, 89, 92].

thumbnail
Download:
Fig 2. Model calibration.

A) Plateaus to the left show the amount of hematopoietic stem cells x0 (upper plateau) and that for MPN stem cells y0 (lower plateau) whereas the plateaus to the right show the amount of hematopoietic stem cells x0(lower plateau) and MPN mature cells y0 (upper plateau). B) Plateaus to the left show the amount of hematopoietic mature cells x1 (upper plateau) and that for MPN mature cells y1 (lower plateau) whereas the plateaus to the right show the amount of hematopoietic mature cells x1 (lower plateau) and MPN mature cells y1 (upper plateau). The yellow and purple boxes show our data used for calibrating (and validating) the model with further details in S1 Appendix. Yellow boxes show our “no MPN cancer values”, and purple boxes show our “full blown” MPN values in the advanced myelofibrosis stage. Yellow position marker shows the number of hematopoietic stem cells as used by Dingli & Michor [92], and black position markers show the number of cells as used by Gentry et al. [86].

https://doi.org/10.1371/journal.pone.0183620.g002

From the steady state condition we have the number of dead cells to be . We further expect rx>dx0 + ax and dx0 ≪dx1. When allowing for MPN development the healthy state becomes unstable when perturbed by the malignant stem cells. Thus, we expect ry >rx.

In the final stage the in silico patient will have vanishing hematopoietic cells and the MPN cells will approach a stable steady state with a higher amount of MPN cells than normal hematopoietic cells in the healthy steady state. This is accomplished by choosing all the c-values equal in order to keep the model as simple as possible and the number of parameters as few as possible. Likewise, the parsimonious principle suggests dy0 = dx0, ay = ax and Ay = Ax.

The JAK2V617F allele burden has been reported to have median values of 7% (95% CL 2–15%; range 1–39%), 33% (95% CL 20–40%; range 1–92%) and 67% (95% CL 52–95%; range 37–99%) in ET, PV and PMF patients, respectively [19]. It follows that the model output perfectly resamples these dynamic changes in the JAK2V617F mutational load (Fig 3). Additional details are given in the S1 Appendix section.

thumbnail
Download:
Fig 3.

Left: Typical development in stem cells (top panel A) and mature cells (bottom panel B). Healthy hematopoietic cells (full blue curves) dominate in the early phase where the number of malignant cells (stipulated red curves) are few. The total number of cells is also shown (dotted green curves). When a stem cell mutates without repairing mechanisms, a slowly increasing exponential growth starts. At a certain stage, the malignant cells become dominant, and the healthy hematopoietic cells begin to show a visible decline. Finally, the composition between the cell types results in a takeover by the malignant cells, leading to an exponential decline in hematopoietic cells and ultimately their extinction. The development is driven by an approximately exponential increase in the MPN stem cells, and the development is closely followed by the mature MPN cells. Right: B)The corresponding allele burden (7%, 33% and 67% corresponding to ET, PV, and PMF, respectively) defined as the ratio of MPN mature cells to the total number of mature cells.

https://doi.org/10.1371/journal.pone.0183620.g003

All these attempts in calibrating the model may simultaneously be considered as validation since they performed successfully. However, the model may be validated further by predicting affected cytokine levels from the inflammatory level. As indicators of the inflammatory level, we refer to those cytokines, which are considered most important in the context of MPNs: IL-1β, IL-1RA, IL-2R, IL-6, IL-8, IL-10, IL-12 and the inflammation biomarker CRP which all turned out to be linearly correlated with the inflammatory level (s).

For the specific cytokines (Ci) tabulated in the S1 Appendix, we have ‘Normal’, ‘PV’, and ‘PMF’ median values (mij, where index i specifies the cytokine and index j refers to ‘Normal’, ‘PV’ and ‘PMF’ states) for each. Then we find ki1 and ki2 such that mij = ki1 sj+ki2 where sj is the value of s at year tj. Similarly, LDH values were demonstrated to be correlated and compared to the total rate of dying cells DI = dx0x0 + dx1x1 + dy0y0 + dy1y1. The results are summarized in Fig 4 which shows that the model predicts data very well. Only IL-6 seems to be less well predicted.

thumbnail
Download:
Fig 4. Model validation.

Cytokines A) IL-1β, B) IL-1RA, C) IL-2R, D) IL-6, E) IL-8, F) IL-10, G) IL-12 and H) C-reactive proteins (CRP) are approximatively linearly correlated with the inflammatory level s. For the specific cytokines, we have from left to right ‘Normal’, ‘PV’, and ‘PMF’ median values (yellow columns) for comparison based on the predicted inflammatory level s (full blue curve) as a function of time after the first insult. I) Similarly, LDH is correlated with and compared to the total rate of dying cells DI = dx0x0 + dx1x1 + dy0y0 + dy1y1.

https://doi.org/10.1371/journal.pone.0183620.g004

Disallowing potential mutations and having no MPN-stem cells initially forces the model system into a steady state where solutions are all constant after a possible initial transient event. Introducing a mutation probability introduces a fatal malignant state; the higher the mutation probability is the faster the malignant state develops. A typical scenario is shown in Fig 3A along with a curve of the allele burden development (Fig 3B). The Figure depicts both modeling of the development of MPN from normal HSC and the early MPN diseases stages (ET/PV) to the advanced myelofibrosis stage.

Having a continuous mutation rate, it will take 24 years for the disease to develop to an allele burden of 7% (e.g. ET) and after additional four years the allele burden reaches 33% (e.g. transformation of ET to PV) to become 67% (e.g. transformation of PV to post-PV myelofibrosis) at year 36 after the first stem cell mutation. Disallowing mutations in the model and initially including a single malignant stem cell and no malignant mature cells shifts the allele burden curve by one year to the left on the time axis.

Thus, the mutation of an HSC to MPN SC triggers the disease. Once an MPN stem cell is established the disease can progress without further mutations.

The baseline inflammatory load (stimulus) is arbitrarily set to 7 pg/ml per day during normal circumstances. It is an exogenous stimulation of the immune system, which leads to an inflammatory level of 3.61 pg/ml, increasing to 3.66 pg/ml in MPNs. This corresponds to a baseline of 700 dead cells (in the hematopoietic steady state) before MPN develops remarkably. A doubling of the baseline inflammatory level is directly affecting the inflammation load (cytokine level) and thereby affecting the rest of the system as dictated by the model equations. In Fig 5 is depicted that shortening the exposure time of inflammation load is associated with deceleration of disease progression.

thumbnail
Download:
Fig 5. Investigation of increased inflammatory load at various onsets and durations.

Blue curve is default parameters corresponding to Fig 3, red dotted is a doubling of inflammatory load, full red curve is a doubling of inflammatory load in year 0–20, then back to default level, black dotted curve is inflammatory doubling from year 10, the full black is inflammatory doubling year 10–30. Upper: Increasing inflammatory load has a boosting effect on MPN MC (A) as well as on HMC (B). Lower: Displaying the results in terms of the clinically available quantity, total blood cell count, also shows a boosted effect with increasing inflammatory load (C). The allele burden of JAK2 mutated blood cells similarly shows that increased inflammation increases disease development (D). There is a clear effect of MPN promotion with increasing inflammatory load, earlier onset, and exposure time. Lowering inflammatory load makes disease progression less rapid. Maintaining a doubling (red dotted curve) shifts the allele burden curve to the left by two years. Shortening the exposure time of inflammatory load weakens the disease progression. The inflammation has a fast impact on the total number of blood cells, which typically changes by 25% within the first year after doubling or reducing the inflammatory load by 50%.

https://doi.org/10.1371/journal.pone.0183620.g005

Discussion

Chronic inflammation is characterized by persistently activated immune cells, DNA damage, tissue destruction, remodeling and fibrosis [106]. In patients with MPNs, these processes are exemplified by the advanced myelofibrosis stage [4, 29], which accordingly might be considered to develop as the consequence of chronic inflammation in the bone marrow–“the inflamed bone marrow” and “the wound that won’t heal” [4, 29,107,108]. Herein, we for the first time use mathematical modelling to substantiate the concept that MPN progression is facilitated by chronic inflammation and that ET and PV are linked through increasing JAK2V617F allele burden [19] which is destined to happen as time increases without interference. Importantly, we were able to create the inflammation-MPN model based upon current knowledge on the interactions between inflammatory cytokines, hematopoietic stem cells and progenitors, and the bone marrow microenvironment [31–33,35–37,105]. By mathematical modelling of all these interactions, our integrated inflammation-MPN model was created. The model was validated from current data on circulating inflammatory cytokines in MPNs [40,44–46], thereby substantiating inflammation to be a highly potent stimulus for clonal evolution and cancer progression in MPNs. In the context that elevated CRP levels have been shown to be associated with shortened leukemia-free survival in myelofibrosis [42], it is of interest that our model was excellently validated by data on CRP levels in the different MPN disease stages as well.

Mathematical modelling has been used to describe the impact of chronic inflammation and immune deregulation in aging [109] and several diseases, including type 1 diabetes mellitus [110], rheumatoid arthritis [96] and colitis-associated colon cancer [111]. Based upon the known association between respiratory infections and chronic inflammation, Herald described a general model of inflammation [97]. In this model, a system of nonlinear ordinary differential equations was used to describe interactions between macrophages, inflammatory and anti-inflammatory cytokines and bacteria. Though initiated by bacteria as the stimulus to trigger chronic inflammation, their study focused on chronic inflammation in the absence of pathogens as well [97]. Of note, even small changes in parameters of importance for inflammatory cytokine production and macrophage sensitivity to cytokines resulted in dramatically different model behaviors [97]. According to this model chronic inflammation is not triggered when the immune system is functioning properly. However, in patients with a dysfunction of the immune system positive feedback of the inflammatory cytokine network is prone to induce chronic inflammation. Furthermore, if the macrophage population is more sensitive to inflammatory cytokines small perturbations initiated by the inflammation stimulus will also lead to chronic inflammation [97]. In this context, it is intriguing to consider if the inherited genetic predisposition to acquire the JAK2V617F-mutation due to the haplotype 46/1 [112–117], which also confers an increased risk of (other) inflammatory diseases (e.g. Crohns’ disease) [118,119] and /or acquired genetic instability due to sustained chronic inflammation (chronic inflammatory diseases or toxin exposure (e.g. smoking) might further increase the risk of developing MPN—a hypothesis originally proposed by Hermouet et al [33,35]. Importantly, the hypersensitivity of clonal MPN-cells to exogenous and endogenous growth factors and inflammatory cytokines might also more easily lead to a chronic inflammatory state–similar to the increased sensitivity of the macrophage population leading to chronic inflammation in the Herald model and also implemented in the Hermouet model, implying an enhanced myelomonocytic response to cytokine stimulation [33,35].

In the Herald model and the model described by Nielsen et al in regard to type 1 diabetes mellitus, the macrophages constituted an important compartment [97,110]. The monocyte-macrophage cell lineage is of major importance in the context of inflammation and cancer development. In our MPN-inflammation model bone marrow macrophages are also of utmost importance—both in regard to release of inflammatory cytokines, but also in regard to the development of myelofibrosis. Thus, in MPNs the monocyte-macrophage cell—together with the megakaryocyte (MK) cell lineage—are considered to be responsible for the development of myelofibrosis by the release of a number of growth factors and inflammatory cytokines that stimulate fibroblast proliferation [36,120,121]. The “Herald Model” is in several aspects equivalent to our model when considering substituting “bacteria”in the “Herald Model” by any noxious inflammatory stimulus. In fact, we implement yet another cell lineage—the MKs—as the source of a continuous release of products that stimulate the vicious inflammation circle, implying ultimately the development of cancer—the MPNs. As previously outlined, our mathematical modelling of the concept of chronic inflammation in MPNs is also supported by the elegant model described by Hermouet and co-workers [33,35], in which the JAK2 46/1 haplotype was proposed as a marker of inappropriate myelomonocytic response to cytokine stimulation, leading to increased risk of inflammation, myeloid neoplasms, and impaired defense against infection [33]. Indeed, the Hermouet model for chronic inflammation [33,35] fits exceedingly well with the Herald model of general inflammation [97] and our mathematical modelling of MPNs as “A Human Inflammation Mode for Cancer Development [29–32]. In this regard, chronic inflammation and immune deregulation in MPNs might act as a trigger for later development of AML and MDS in line with the known association of inflammatory signaling and cancer [24–27]. The above models are additionally supported by the hypothetical model by Takizawa et al. (2010) [122], describing how chronic inflammatory processes might impinge on hematopoiesis, potentially fostering hematopoietic stem cell diseases, including MPNs. By inducing high proliferation of most HSCs, chronic inflammation might give rise to both exhaustion of the HSC pool and an even greater risk to accumulate genetic alterations in HSCs. Furthermore, by inflammatory stimuli from the bone marrow microenvironment these genetically altered HSCs might be rescued or “cancer cell niche” for later development of a hematological cancer [122].

The perspectives of our study are several. In the context that myelomonocytic cells (granulocytes, macrophages, monocytes) and MKs are all deeply involved in cancer development and progression [123,124], chronic inflammation is associated with premature atherosclerosis (atherothrombosis) [21–23, 29,30], in which both platelets and monocytes are highly important (monocytes a link between atherosclerosis and cancer [28]) and platelets are intimately involved in the metastatic process in cancer [124]—and likely in MPNs as well [125]—the avenue is opened for studying all these aspects by using mathematical modelling of current knowledge of the impact of chronic inflammation and immune deregulation in patients with MPNs. Ultimately, mathematical modelling may also be able to substantiate which agents to be used in MPNs in order to induce “minimal residual disease”[125–129] and the importance of early intervention with agents that directly target both the malignant clone (interferon-alpha2) [126–129] and the inflammatory process (JAK1-2 inhibition with e.g. ruxolitinib) [130].

In conclusion, we have for the first time applied mathematical modelling as a tool to deliver the proof of concept that chronic inflammation is closely linked to the development of the MPNs—myeloproliferative cancers which today are considered to be “chronic inflammatory diseases”, in which chronic inflammation may be a driving force for clonal expansion and ultimately the development of AML [4, 29–32,39]. Studies are ongoing to elucidate the above perspectives by mathematical modelling. In this regard, mathematical modelling of resolution of inflammation may be highly important [98] and useful to support the decision-making which agents to use in the future for patients with MPNs in order to induce minimal residual disease and hopefully cure.

References

  1. 1. Campbell PJ, Green AR. The myeloproliferative disorders. N Engl J Med. 2006; 355 (23): 2452–2466. pmid:17151367
  2. 2. Marchioli R, Finazzi G, Landolfi R, Kutti J, Gisslinger H, Patrono C, et al. Vascular and neoplastic risk in a large cohort of patients with polycythemia vera. J Clin Oncol. 2005; 23: 2224–2232. pmid:15710945
  3. 3. Kristinsson SY, Landgren O, Samuelsson J Björkholm M, Goldin LR. Autoimmunity and the risk of myeloproliferative neoplasms. Haematologica. 2010; 95(7): 1216–1220. pmid:20053870
  4. 4. Hasselbalch HC, Bjørn ME. MPNs as Inflammatory Diseases: The Evidence, Consequences, and Perspectives. Mediators Inflamm. 2015: 102476. pmid:26604428
  5. 5. Frederiksen H, Farkas DK, Christiansen CF Hasselbalch HC, Sørensen HT. Chronic myeloproliferative neoplasms and subsequent cancer risk: a Danish population-based cohort study. Blood. 2011; 118(25): 6515–6520. pmid:22039256
  6. 6. Frederiksen H, Farkas DK, Christiansen CF, Larsen TS, Hasselbalch HC, Stentoft J, et al. Survival of patients with chronic myeloproliferative neoplasms and new primary cancers: a population-based cohort study. Lancet Haematol. 2015; 2(7): e289–296. pmid:26688384
  7. 7. Pettersson H, Knutsen H, Holmberg E, Andreasson B. Increased incidence of another cancer in myeloproliferative neoplasms patients at the time of diagnosis. Eur J Haematol. 2015; 94(2): 152–156. pmid:25039361
  8. 8. Enblom A, Lindskog E, Hasselbalch HC, Hersby D, Bak M, Tetu J, et al. High rate of abnormal blood values and vascular complications before diagnosis of myeloproliferative neoplasms. Eur J Intern Med. 2015; 26(5): 344–347. pmid:25863408
  9. 9. Sørensen AL, Hasselbalch HC. Antecedent cardiovascular disease and autoimmunity in Philadelphia-negative chronic myeloproliferative neoplasms. Leuk Res. 2016; 41: 27–35. pmid:26718091
  10. 10. Bjorkholm M, Derolf AR, Hultcrantz M, Kristinsson SY, Ekstrand C, Goldin LR. Treatment-related risk factors for transformation to acute myeloid leukemia and myelodysplastic syndromes in myeloproliferative neoplasms. J Clin Oncol. 2011: 29(17): 2410–2415. pmid:21537037
  11. 11. James C, Ugo V, Le Couédic J-P, Staerk J, Delhommeau F, Lacout C, et al. A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera. Nature. 2005; 434: 1144–1148. pmid:15793561
  12. 12. Kralovics R, Passamonti F, Buser AS, Teo SS, Tiedt R, Passweg JR, et al. A gain-of-function mutation of JAK2 in myeloproliferative disorders. N Engl J Med. 2005; 352(17): 1779–1790. pmid:15858187
  13. 13. Baxter EJ, Scott LM, Campbell PJ, East C, Fourouclas N, Swanton S, et al. Acquired mutation of the tyrosine kinase JAK2 in human myeloproliferative disorders. Lancet. 2005; 365(9464): 1054–1061.
  14. 14. Levine RL, Wadleigh M, Cools J, Ebert BL, Wernig G, Huntly BJ, et al. Activating mutation in the tyrosine kinase JAK2 in polycythemia vera, essential thrombocythemia, and myeloid metaplasia with myelofibrosis. Cancer Cell. 2005; 7(4): 387–397. pmid:15837627
  15. 15. Klampfl T, Gisslinger H, Harutyunyan AS, et al. Somatic mutations of calreticulin in myeloproliferative neoplasms. N Engl J Med. 2013; 369(25): 2379–2390. pmid:24325356
  16. 16. Nangalia J, Massie CE, Baxter EJ, Nivarthi H, Rumi E, Milosevic JD, et al. Somatic CALR mutations in myeloproliferative neoplasms with nonmutated JAK2. N Engl J Med. 2013; 369(25): 2391–2405. pmid:24325359
  17. 17. Cazzola M, Kralovics R. From Janus kinase 2 to calreticulin: The clinically relevant genomic landscape of myeloproliferative neoplasms. Blood. 2014; 123(24): 3714–3719. pmid:24786775
  18. 18. Lundberg P, Karow A, Nienhold R, Looser R, Hao-Shen H, Nissen I, et al. Clonal evolution and clinical correlates of somatic mutations in myeloproliferative neoplasms. Blood. 2014; 123(14): 2220–2228. pmid:24478400
  19. 19. Larsen TS, Pallisgaard N, Møller MB, Hasselbalch HC. The JAK2 V617F allele burden in essential thrombocythemia, polycythemia vera and primary myelofibrosis—impact on disease phenotype. Eur J Haematol. 2007; 79(6); 508–515. pmid:17961178
  20. 20. Barosi G, Rosti V, Bonetti E, Campanelli R, Carolei A, Catarsi P, et al. Evidence that Prefibrotic Myelofibrosis Is Aligned along a Clinical and Biological Continuum Featuring Primary Myelofibrosis. PLoS One. 2012; 7(4): e35631. pmid:22536419
  21. 21. Danesh J, Whincup P, Walker M, Lennon L, Thomson A, Appleby P, et al. Low grade inflammation and coronary heart disease: prospective study and updated meta-analyses. BMJ. 2000; 321: 199–204. pmid:10903648
  22. 22. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Plasma concentration of C-reactive protein and risk of developing peripheral vascular disease. Circulation. 1998; 97(5): 425–428. pmid:9490235
  23. 23. Brevetti G, Giugliano G, Brevetti L, Hiatt WR. Inflammation in peripheral arterial disease. Circulation. 2010; 122: 1862–1875. pmid:21041698
  24. 24. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000; 100(1): 57–70. pmid:10647931
  25. 25. Pikarsky E, Porat RM, Stein I, Abramovitch R, Amit S, Kasem S, et al. NF-kappaB functions as a tumour promoter in inflammation associated cancer. Nature. 2004; 431(7007): 461–466. pmid:15329734
  26. 26. Karin M, Greten FR. NF-kappaB: linking inflammation and immunity to cancer development and progression. Nat Rev Immunol. 2005: 5(10): 749–759. pmid:16175180
  27. 27. Mantovani A, Garlanda C, Allavena P. Molecular pathways and targets in cancer-related inflammation. Ann Med. 2010; 42(3): 161–170. pmid:20384432
  28. 28. Pittet MJ, Swirski FK. Monocytes link atherosclerosis and cancer. Eur J Immunol. 2011. 41(9): 2519–2522. pmid:21952809
  29. 29. Hasselbalch HC. Chronic inflammation as a promotor of mutagenesis in essential thrombocythemia, polycythemia vera and myelofibrosis. A human inflammation model for cancer development? Leuk Res. 2013; 37(2): 214–220. pmid:23174192
  30. 30. Hasselbalch HC. Perspectives on chronic inflammation in essential thrombocythemia, polycythemia vera, and myelofibrosis: is chronic inflammation a trigger and driver of clonal evolution and development of accelerated atherosclerosis and second cancer? Blood. 2012; 119: 3219–3225. pmid:22318201
  31. 31. Hasselbalch HC. The role of cytokines in the initiation and progression of myelofibrosis. Cytokine Growth Factor Rev. 2013; 24(2): 133–145. pmid:23415024
  32. 32. Hasselbalch HC. A role of NF-E2 in chronic inflammation and clonal evolution in essential thrombocythemia, polycythemia vera and myelofibrosis? Leuk Res. 2014; 38(2): 263–266. pmid:23932394
  33. 33. Hermouet S, Vilaine M. The JAK2 46/1 haplotype: a marker of inappropriate myelomonocytic response to cytokine stimulation, leading to increased risk of inflammation, myeloid neoplasm, and impaired defense against infection? Haematologica. 2011; 96(11): 1575–1579. pmid:22058280
  34. 34. Hermouet S, Hasselbalch HC, Čokić V. Mediators of Inflammation in Myeloproliferative Neoplasms: State of the Art. Mediators Inflamm. 2015; 964613. pmid:26681841
  35. 35. Hermouet S, Bigot-Corbel E, Gardie B. Pathogenesis of Myeloproliferative Neoplasms: Role and Mechanisms of Chronic Inflammation. Mediators Inflamm. 2015; 145293. pmid:26538820
  36. 36. Desterke C, Martinaud C, Ruzehaji N, Le Bousse-Kerdilès MC. Inflammation as a Keystone of Bone Marrow Stroma Alterations in Primary Myelofibrosis. Mediators Inflamm. 2015; 415024. pmid:26640324
  37. 37. Čokić VP, Mitrović-Ajtić O, Beleslin-Čokić BB, Marković D, Buač M, Diklić M, et al. Proinflammatory Cytokine IL-6 and JAK-STAT Signaling Pathway in Myeloproliferative Neoplasms. Mediators Inflamm. 2015; 453020. pmid:26491227
  38. 38. Geyer HL, Dueck AC, Scherber RM, Mesa R. Impact of Inflammation on Myeloproliferative Neoplasm Symptom Development. Mediators Inflamm. 2015; 284706. pmid:26538823
  39. 39. Koschmieder S, Mughal TI, Hasselbalch HC, Barosi G, Valent P, Kiladjian JJ, et al. Myeloproliferative neoplasms and inflammation: whether to target the malignant clone or the inflammatory process or both. Leukemia. 2016; 30(5): 1018–1024. pmid:26854026
  40. 40. Panteli KE, Hatzimichael EC, Bouranta PK, Katsaraki A, Seferiadis K, Stebbing , et al. Serum interleukin (IL)-1, IL-2, sIL-2Ra, IL-6 and thrombopoietin levels in patients with chronic myeloproliferative diseases. Br J Haematol. 2005; 130: 709–715. pmid:16115126
  41. 41. Barbui T, Carobbio A, Finazzi G, Vannucchi AM, Barosi G, Antonioli E et al. Inflammation and thrombosis in essential thrombocythemia and polycythemia vera: different role of C-reactive protein and pentraxin 3. Haematologica. 2011; 96: 315–318. pmid:21173097
  42. 42. Barbui T, Carobbio A, Finazzi G, Guglielmelli P, Salmoiraghi S, Rosti V et al., Elevated C-reactive protein is associated with shortened leukemia-free survival in patients with myelofibrosis. Leukemia. 2013; 27(10): 2084–6. pmid:23828261
  43. 43. Verstovsek S, Kantarjian H, Mesa RA, Pardanani AD, Cortes-Franco J, Thomas DA, et al. Safety and efficacy of INCB018424, a JAK1 and JAK2 inhibitor, in myelofibrosis. N Engl J Med. 2010; 363: 1117–1127. pmid:20843246
  44. 44. Tefferi A, Vaidya R, Caramazza D, Finke C, Lasho T, Pardanani A. Circulating interleukin (IL)-8, IL-2R, IL-12, and IL-15 levels are independently prognostic in primary myelofibrosis: a comprehensive cytokine profiling study. J Clin Oncol 29. 2011; 1356–1363. pmid:21300928
  45. 45. Vaidya R, Gangat N, Jimma T, Finke CM, Lasho TL, Pardanani A, et al. Plasma cytokines in polycythemia vera: phenotypic correlates, prognostic relevance, and comparison with myelofibrosis. Am J Hematol. 2012; 87(11): 1003–5. pmid:22965887
  46. 46. Pourcelot E, Trocme C, Mondet J, Bailly S, Toussaint B, Mossuz P. Cytokine profiles in polycythemia vera and essential thrombocythemia patients: Clinical implications. Experimental Hematol. 2014; 42: 360–368.
  47. 47. Christensen AS, Møller JB, Hasselbalch HC. Chronic kidney disease in patients with the Philadelphia-negative chronic myeloproliferative neoplasms. Leuk Res. 2014; 38(4): 490–495. pmid:24630365
  48. 48. Bak M, Sørensen TL, Meulengracht Flachs E, Zwisler AD, Juel K, Frederiksen H, et al: Age-Related Macular Degeneration in Patients with Chronic Myeloproliferative Neoplasms: A Nationwide Population-Based Cohort Study. JAMA Ophthalmol. 2017; 135(8):835–843. pmid:28655032
  49. 49. Hasselbalch HC. Smoking as a contributing factor for development of polycythemia vera and related neoplasms. Leuk Res. 2015; 015 Sep 9.
  50. 50. Lindholm Sørensen A, Hasselbalch HC. Smoking and Philadelphia-negative chronic myeloproliferative neoplasms. Eur J Haematol. 2016; 97(1): 63–69. pmid:26384085
  51. 51. Andersen CL, Bjørn ME, McMullin MF, Harrison C, Samuelsson J,Ejerblad E et al. Circulating YKL-40 in patients with essential thrombocythemia and polycythemia vera treated with the novel histone deacetylase inhibitor vorinostat. Leuk Res. 2014; 38(7): 816–821. pmid:24836761
  52. 52. Bjørn ME, Andersen CL, Jensen MK, Hasselbalch HC. Circulating YKL-40 in myelofibrosis a potential novel biomarker of disease activity and the inflammatory state. Eur J Haematol. 2014; 93(3): 224–228. pmid:24689875
  53. 53. Bjørn ME, Hasselbalch HC. The Role of Reactive Oxygen Species in Myelofibrosis and Related Neoplasms. Mediators Inflamm. 2015; 648090. pmid:26538833
  54. 54. Fleischman AG, Aichberger KJ, Luty SB, Bumm TG, Petersen CL, Doratotaj S, et al: TNFα facilitates clonal expansion of JAK2V617F positive cells in myeloproliferative neoplasms. Blood. 2011; 118: 6392–6398. pmid:21860020
  55. 55. Lu M, Xia L, Liu YC, Hochman T et al: Lipocalin produced by myelofibrosis cells affects the fate of both hematopoietic and marrow microenvironmental cells. Blood. 2015; 126: 972–982. pmid:26022238
  56. 56. Kleppe M, Kwak M, Koppikar P, Riester M, Keller M, Bastian L et al. JAK-STAT pathway activation in malignant and nonmalignant cells contributes to MPN pathogenesis and therapeutic response. Cancer Discov. 2015; 5: 316–331. pmid:25572172
  57. 57. Kaufmann KB, Grunder A, Hadlich T, Wehrle J, Gothwal M, Bogeska R, et al: A novel murine model of myeloproliferative disorders generated by overexpression of the transcription factor NF-E2. J Exp Med. 2012; 209: 35–50. pmid:22231305
  58. 58. Jutzi JS, Bogeska R, Nikoloski G, Schmid CA, Seeger TS, Stegelmann F et al: MPN patients harbor recurrent truncating mutations in transcription factor NF-E2. J Exp Med. 2013; 210: 1003–1019. pmid:23589569
  59. 59. Jutzi JS, Pahl HL. The hen or the egg: inflammatory aspects of murine MPN models. Mediators Inflamm. 2015; 2015:101987. pmid:26543325
  60. 60. Wehrle J, Seeger TS, Schwemmers S, Pfeifer D, Bulashevska A, Pahl HL. Transcription factor nuclear factor erythroid-2 mediates expression of the cytokine interleukin 8, a known predictor of inferior outcome in patients with myeloproliferative neoplasms. Haematologica. 2013; 98: 1073–1080. pmid:23445878
  61. 61. Roelz R, Pilz IH, Mutschler M, Pahl HL. Of mice and men: human RNA polymerase III promoter U6 is more efficient than its murine homologue for shRNA expression from a lentiviral vector in both human and murine progenitor cells. Exp Hematol. 2010; 38: 792–797. pmid:20685233
  62. 62. Zhang Y, Liu X, McHale C, Li R, Zhang L, Wu Y et al. Bone marrow injury induced via oxidative stress in mice by inhalation exposure to formaldehyde. PLoS One. 2013; 8: e74974. pmid:24040369
  63. 63. Zhang Q, Zhao K, Shen Q, Han Y, Gu Y, Li X, et al: Tet2 is required to resolve inflammation by recruiting Hdac2 to specifically repress IL-6. Nature. 2015; 525: 389–393. pmid:26287468
  64. 64. Skov V, Larsen TS, Thomassen M, Riley CH, Jensen MK, Bjerrum OW, et al: Whole-blood transcriptional profiling of interferon-inducible genes identifies highly upregulated IFI27 in primary myelofibrosis. Eur J Haematol. 2011; 87(1): 54–60. pmid:21447007
  65. 65. Skov V, Riley CH, Thomassen M, Larsen TS, Jensen MK, Bjerrum OW, et al: Whole blood transcriptional profiling reveals significant down-regulation of human leukocyte antigen class I and II genes in essential thrombocythemia, polycythemia vera and myelofibrosis. Leuk Lymphoma. 2013; 54(10): 2269–2273. pmid:23302045
  66. 66. Skov V, Thomassen M, Riley CH, Jensen MK, Bjerrum OW, Kruse TA, et al: Gene expression profiling with principal component analysis depicts the biological continuum from essential thrombocythemia over polycythemia vera to myelofibrosis. Exp. Hematol. 2012; 40(9): 771–780. pmid:22659388
  67. 67. Skov V, Larsen TS, Thomassen M, Riley CH, Jensen MK, Bjerrum OW, et al. Molecular profiling of peripheral blood cells from patients with polycythemia vera and related neoplasms: identification of deregulated genes of significance for inflammation and immune surveillance. Leuk Res. 2012; 36(11): 1387–1392. pmid:22877729
  68. 68. Hasselbalch HC, Thomassen M, Riley CH, Kjær L, Larsen TS, Jensen MK, et al. Whole blood transcriptional profiling reveals deregulation of oxidative and antioxidative defence genes in myelofibrosis and related neoplasms. Potential implications of downregulation of Nrf2 for genomic instability and disease progression. PLoS One. 2014; 14;9(11):e112786. pmid:25397683
  69. 69. Skov V, Burton M, Thomassen M, Larsen TS, Riley CH, Madelung AB et al. A 7-Gene Signature Depicts The Biochemical Profile of Early Prefibrotic Myelofibrosis. Plos One. 2016; 11(8): e0161570. pmid:27579896
  70. 70. Ciaffoni F, Cassella E, Varricchio L, Massa M, Barosi G, Migliaccio AR. Activation of non-canonical TGF-β1 signaling indicates an autoimmune mechanism for bone marrow fibrosis in primary myelofibrosis. Blood Cells Mol Dis. 2015; 54(3): 234–41 pmid:25703685
  71. 71. Riley CH, Jensen MK, Brimnes MK, Hasselbalch HC, Bjerrum OW, Straten PT, et al. Increase in circulating CD4(+) CD25(+) Foxp3(+) T cells in patients with Philadelphia-negative chronic myeloproliferative neoplasms during treatment with IFN-alpha. Blood. 2011; 118(8): 2170–2173. pmid:21708889
  72. 72. Riley CH, Hansen M, Brimnes MK, Hasselbalch HC, Bjerrum OW, Svane IM, et al. Expansion of circulating CD56bright natural killer cells in patients with JAK2-positive chronic myeloproliferative neoplasms during treatment with interferon-α. Eur J Haematol. 2015; 94(3): 227–234. pmid:25082025
  73. 73. Riley CH, Brimnes MK, Hansen M, Jensen MK, Hasselbalch HC, Kjaer L, et al: Interferon-alpha induces marked alterations in circulating regulatory T cells, NK cell subsets and dendritic cells in patients with JAK2 -positive essential thrombocythemia and polycythemia vera. Eur J Haematol. 2016; 97(1): 83–92. pmid:26385526
  74. 74. Barosi G. An immune dysregulation in MPN. Curr Hematol Malig Rep. 2014; 9(4): 331–339. pmid:25139710
  75. 75. Titmarsh GJ, McMullin MF, McShane CM, Clarke M, Engels EA, Anderson LA, et al: Community-acquired infections and their association with myeloid malignancies. Cancer Epidemiol. 2014; 38(1): 56–61. pmid:24275260
  76. 76. Nielsen C, Birgens HS, Nordestgaard BG, Kjaer L, Bojesen SE. The JAK2 V617F somatic mutation, mortality and cancer risk in the general population. Haematologica. 2011; 96(3): 450–453. pmid:21160067
  77. 77. Nielsen C, Bojesen SE, Nordestgaard BG, Kofoed KF, Birgens HS. JAK2V617F somatic mutation in the general population: myeloproliferative neoplasm development and progression rate. Haematologica. 2014; 99(9): 1448–1455. pmid:24907356
  78. 78. Haeno H, Levine RL, Gilliland DG, Michor F. A progenitor cell origin of myeloid malignancies. Proc Natl Acad Sci U S A. 2009; 106(39): 16616–16621. pmid:19805346
  79. 79. Attolini CS, Cheng YK, Beroukhim R, Getz G, Abdel-Wahab O, Levine RL, et al. A mathematical framework to determine the temporal sequence of somatic genetic events in cancer. Proc Natl Acad Sci U S A. 2010; 107(41): 17604–17609. pmid:20864632
  80. 80. Michor F, Iwasa Y, Nowak MA. Dynamics of cancer progression. Nature Reviews Cancer. 2004; 4: 197–205. pmid:14993901
  81. 81. Michor F, Hughes TP, Iwasa Y, Branford S, Shah NP, Sawyers CL. Dynamics of chronic myeloid leukaemia. Nature. 2005; 435: 1267–1270. pmid:15988530
  82. 82. De Pillis LG, Radunskaya AE, Wiseman CL. A validated mathematical model of cell-mediated immune response to tumor growth. Cancer Research. 2005; 65(17): 7950–7958. pmid:16140967
  83. 83. Michor F. Mathematical models of cancer stem cells. J Clin Oncol. 2008; 26(17): 2854–2861. pmid:18539964
  84. 84. Stiehl T, Marciniak-Czochra A. Characterization of stem cells using mathematical models of multistage cell lineages. Mathematical and Computer Modelling. 2011; 53(7–8): 1505–1517.
  85. 85. Stiehl T, Marciniak-Czochra A. Mathematical Modeling of Leukemogenesis and Cancer Stem Cell Dynamics. Math Model Nat Phenom. 2012; 7(1): 166–202.
  86. 86. Gentry SN, Jackson TL. A mathematical model of cancer stem cell driven tumor initiation: implications of niche size and loss of homeostatic regulatory mechanisms. Plos One. 2013; 8(8): e71128. pmid:23990931
  87. 87. Walenda T, Stiehl T, Braun H, Fröbel J, Ho AD, Schroeder T, et al: Feedback Signals in Myelodysplastic Syndromes: Increased Self-Renewal of the Malignant Clone Suppresses Normal Hematopoiesis. PLoS Comput Biol. 2014; 10(4): e1003599. pmid:24763223
  88. 88. Stiehl T, Baran N, Ho AD, Marciniak-Czochra A. Cell Division Patterns in Acute Myeloid Leukemia Stem-like Cells Determine Clinical Course: A Model to Predict Patient Survival. Cancer Research. 2015; 75(6): 940–994. pmid:25614516
  89. 89. Stiehl T, Lutz C, Marciniak-Czochra A. Emergence of heterogeneity in acute leukemias. Biology Direct. 2016; 11: 51. pmid:27733173
  90. 90. Goldie JH, Coldman AJ. A mathematic model for relating the drug sensitivity of tumors to their spontaneous mutation rate. Cancer Treatment Reports. 1979; 63: 1727–1733. pmid:526911
  91. 91. Goldie JH, Coldman AJ. Quantitative model for multiple levels of drug resistance in clinical tumors. Cancer Treatment Reports. 1983; 67: 923–931. pmid:6627236
  92. 92. Dingli D, Michor F. Successful therapy must eradicate cancer stem cells. Stem Cells. 2006; 24(12): 2603–2610. pmid:16931775
  93. 93. De Pillis LG and Radunskaya A: The dynamics of an optimal controlled tumor model: a case study. Mathematical and computer modelling. 2003; 37: 1221–1244.
  94. 94. Saleem M, Agrawal T: Complex dynamics in a mathematical model of tumor growth with time delays in cell proliferation. International Journal of Scientific and Research Publications. 2012; 2(6): 1–7.
  95. 95. Nielsen KH, Pociot FM, Ottesen JT. Bifurcation Analysis of an Existing Mathematical Model Reveals Novel Treatment Strategies and Suggests Potential Cure for Type 1 Diabetes. Mathematical Medicine and Biology. 2014; 31(3): 205–225. pmid:23620354
  96. 96. Baker M, Denman-Johnson S, Brook BS, Gaywood I, Owen MR. Mathematical modelling of cytokine-mediated inflammation in rheumatoid arthritis. Mathematical Medicine and Biology. 2013. 30: 311–337. pmid:23002057
  97. 97. Herald MC. General Model of Inflammation. Bulletin of Mathematical Biology. 2010; 72: 765–779. pmid:19851812
  98. 98. Dunster JL, Byrne HM, King JR. The Resolution of Inflammation: A Mathematical Model of Neutrophil and Macrophage Interactions. Bull Math Biol. 2014; 76: 1953–1980. pmid:25053556
  99. 99. Jacquez JA, Simon CP. Qualitative Theory of Compartmental Systems. SIAM Review. 1993; 35(1): 43–79.
  100. 100. Michor F, Iwasa Y, Nowak MA. The age incidence of chronic myeloid leukemia can be explained by a one-mutation model. Proc Natl Acad Sci USA. 2006; 103: 14931–14934. pmid:17001000
  101. 101. Jackson AL, Loeb LA. The mutation rate and cancer. Genetics. 1998; 148: 1483–1490. pmid:9560368
  102. 102. Cooke MS, Evans MD, Dizdaroglu M, Lunec J. Oxidative DNA damage: mechanisms, mutation, and disease. FASEB J. 2003; 17: 1195–1214. pmid:12832285
  103. 103. Ferguson LR. Chronic inflammation and mutagenesis. Mutat Res. 2010; 690: 3–11. pmid:20223251
  104. 104. Kiraly O, Gong G, Olipitz W, Muthupalani S, Engelward BP. Inflammation-Induced Cell Proliferation Potentiates DNA Damage-Induced Mutations in Vivo. Plos Genet. 2015; 11(2):e100490.
  105. 105. King KY, Goodell MA. Inflammatory modulation of hematopoietic stem cells: viewing the hematopoietc stem cell as a foundation for the immune response. Nat Rev Immmunol. 2014; 11(10): 685–692.
  106. 106. Rockey DC, Bell PD, Hill JA. Fibrosis—A Common Pathway to Organ Injury and Fibrosis. N Engl J Med. 2015; 372: 1138–1149. pmid:25785971
  107. 107. Dvorak HF. Tumors: wounds that do not heal, Similarities between tumor stroma generation and wound healing. N Engl J Med. 1986; 315(26): 1650–1659. pmid:3537791
  108. 108. Dvorak HF. Tumors: wounds that do not heal-redux. Cancer Immunology Research. 2015; 3(1): 1–11. pmid:25568067
  109. 109. Nikas JB. Inflammation and Immune System Activation in Aging: A Mathematical Approach. Scientific Reports. 2013; 3: 3254: 1–7. pmid:24247109
  110. 110. Nielsen KH, Pociot FM, Ottesen JT. Bifurcation analysis of an existing mathematical model reveals novel treatment strategies and suggests potential cure for type 1 diabetes. Math Med Biol. 2014; 31(3): 205–225. pmid:23620354
  111. 111. Lo WC, Martin EW Jr, Hitchcock CL, Friedman A. Mathematical model of colitis-associated colon cancer. J Theor Biol. 2013; 317: 20–29. pmid:23026764
  112. 112. Pardanani A, Fridley BL, Lasho TL, Gilliland DG, Tefferi A. Host genetic variation contributes to phenotypic diversity in myeloproliferative disorders. Blood. 2008; 111(5): 2785–2789. pmid:18006699
  113. 113. Kilpivaara O, Mukherjee S, Schram AM, Wadleigh M, Mullally A, Ebert BL, et al. A germline JAK2 SNP is associated with predisposition to the development of JAK2(V617F)-positive myeloproliferative neoplasms. Nat Genet. 2009; 41(4): 455–459. pmid:19287384
  114. 114. Jones AV, Chase A, Silver RT, Oscier D, Zoi K, Wang YL, et al: JAK2 haplotype is a major risk factor for the development of myeloproliferative neoplasms. Nat Genet. 2009; 41(4): 446–449. pmid:19287382
  115. 115. Olcaydu D, Harutyunyan A, Jäger R, Berg T, Gisslinger B, Pabinger I, et al: A common JAK2 haplotype confers susceptibility to myeloproliferative neoplasms. Nat Genet. 2009; 41(4): 450–454. pmid:19287385
  116. 116. Cross NC. Genetic and epigenetic complexity in myeloproliferative neoplasms. Hematology Am Soc Hematol Educ Program. 2011; 208–214. pmid:22160036
  117. 117. Tapper W, Jones AV, Kralovics R, Harutyunyan AS, Zoi K, Leung W, et al: Genetic variation at MECOM, TERT, JAK2 and HBS1L-MYB predisposes to myeloproliferative neoplasms. Nat Commun. 2015; 6: 6691. pmid:25849990
  118. 118. Barrett JC, Hansoul S, Nicolae DL, Cho JH, Duerr RH, Rioux JD, et al: Genome-wide association defines more than 30 distinct susceptibility loci for Crohn's disease. Nat Genet. 2008; 40(8): 955–962. pmid:18587394
  119. 119. Ferguson LR, Han DY, Fraser AG, Huebner C, Lam WJ, Morgan AR, et al: Genetic factors in chronic inflammation: single nucleotide polymorphisms in the STAT-JAK pathway, susceptibility to DNA damage and Crohn's disease in a New Zealand population. Mutat Res. 2010; 690(1–2): 108–115. pmid:20109474
  120. 120. Martyre MC, Romquin N, Le Bousse-Kerdiles MC, Chevillard S, Benyahia B, Dupriez B, et al. Transforming growth factor-beta and megakaryocytes in the pathogenesis of idiopathic myelofibrosis. Br J Haematol. 1994; 88(1): 9–16. pmid:7803262
  121. 121. Tefferi A. Myelofibrosis with myeloid metaplasia. N Engl J Med. 2000; 27; 342(17): 1255–1265. pmid:10781623
  122. 122. Takizawa H, Boettcher S, Manz MG. Demand-adapted regulation of early hematopoiesis in infection and inflammation. Blood. 2012; 119(13): 2991–3002. pmid:22246037
  123. 123. Parihar A, Eubank TD, Doseff AI. Monocytes and macrophages regulate immunity through dynamic networks of survival and cell death. J Innate Immun. 2010; 2: 204–215. pmid:20375558
  124. 124. Leblanc R, Peyruchaud O. Metastasis: new functional implications of platelets and megakaryocytes. Blood. 2016; 128(1): 24–31. pmid:27154188
  125. 125. Hasselbalch HC. The platelet-cancer loop in myeloproliferative cancer. Is thrombocythemia an enhancer of cancer invasiveness and metastasis in essential thrombocythemia, polycythemia vera and myelofibrosis? Leuk Res. 2014; 38(10): 1230–1236. pmid:25149709
  126. 126. Hasselbalch HC. A new era for IFN-α in the treatment of Philadelphia-negative chronic myeloproliferative neoplasms. Expert Rev Hematol. 2011; 4(6): 637–655. pmid:22077528
  127. 127. Silver RT, Kiladjian J-J, Hasselbalch HC. Interferon and the treatment of polycythemia vera, essential thrombocythemia and myelofibrosis. Expert Rev Hematol. 2013; 6(1): 49–58. pmid:23373780
  128. 128. Rank C, Bjerrum O, Larsen T, Kjær L, de Stricker K, Riley CH, et al: Minimal residual disease after long-term interferon-alpha2 treatment: a report on hematological, molecular and histomorphological response patterns in 10 patients with essential thrombocythemia and polycythemia vera. Leuk Lymphoma. 2015; 1–7.
  129. 129. Kiladjian JJ, Giraudier S, Cassinat B. Interferon-alpha for the therapy of myeloproliferative neoplasms: targeting the malignant clone. Leukemia. 2016; 30(4): 776–781. pmid:26601783
  130. 130. Hasselbalch HC. Perspectives on the impact of JAK-inhibitor therapy upon inflammation-mediated comorbidities in myelofibrosis and related neoplasms. Expert Rev Hematol. 2014; 7(2): 203–216. pmid:24524202