The role of TNF-α inhibitor in glioma virotherapy: A mathematical model

  • Received: 22 June 2016 Accepted: 30 June 2016 Published: 01 January 2017
  • MSC : Primary: 92C42, 92C50; Secondary: 49J15

  • Virotherapy, using herpes simplex virus, represents a promising therapy of glioma. But the innate immune response, which includes TNF-α produced by macrophages, reduces the effectiveness of the treatment. Hence treatment with TNF-α inhibitor may increase the effectiveness of the virotherapy. In the present paper we develop a mathematical model that includes continuous infusion of the virus in combination with TNF-α inhibitor. We study the efficacy of the treatment under different combinations of the two drugs for different scenarios of the burst size of newly formed virus emerging from dying infected cancer cells. The model may serve as a first step toward developing an optimal strategy for the treatment of glioma by the combination of TNF-α inhibitor and oncolytic virus injection.

    Citation: Elzbieta Ratajczyk, Urszula Ledzewicz, Maciej Leszczynski, Avner Friedman. The role of TNF-α inhibitor in glioma virotherapy: A mathematical model[J]. Mathematical Biosciences and Engineering, 2017, 14(1): 305-319. doi: 10.3934/mbe.2017020

    Related Papers:

    [1] Gesham Magombedze, Winston Garira, Eddie Mwenje . Modelling the immunopathogenesis of HIV-1 infection and the effect of multidrug therapy: The role of fusion inhibitors in HAART. Mathematical Biosciences and Engineering, 2008, 5(3): 485-504. doi: 10.3934/mbe.2008.5.485
    [2] Joseph Malinzi, Rachid Ouifki, Amina Eladdadi, Delfim F. M. Torres, K. A. Jane White . Enhancement of chemotherapy using oncolytic virotherapy: Mathematical and optimal control analysis. Mathematical Biosciences and Engineering, 2018, 15(6): 1435-1463. doi: 10.3934/mbe.2018066
    [3] Donggu Lee, Aurelio A. de los Reyes V, Yangjin Kim . Optimal strategies of oncolytic virus-bortezomib therapy via the apoptotic, necroptotic, and oncolysis signaling network. Mathematical Biosciences and Engineering, 2024, 21(3): 3876-3909. doi: 10.3934/mbe.2024173
    [4] Yangjin Kim, Seongwon Lee, You-Sun Kim, Sean Lawler, Yong Song Gho, Yoon-Keun Kim, Hyung Ju Hwang . Regulation of Th1/Th2 cells in asthma development: A mathematical model. Mathematical Biosciences and Engineering, 2013, 10(4): 1095-1133. doi: 10.3934/mbe.2013.10.1095
    [5] Nikolay L. Martirosyan, Erica M. Rutter, Wyatt L. Ramey, Eric J. Kostelich, Yang Kuang, Mark C. Preul . Mathematically modeling the biological properties of gliomas: A review. Mathematical Biosciences and Engineering, 2015, 12(4): 879-905. doi: 10.3934/mbe.2015.12.879
    [6] Shengqiang Liu, Lin Wang . Global stability of an HIV-1 model with distributed intracellular delays and a combination therapy. Mathematical Biosciences and Engineering, 2010, 7(3): 675-685. doi: 10.3934/mbe.2010.7.675
    [7] B. M. Adams, H. T. Banks, Hee-Dae Kwon, Hien T. Tran . Dynamic Multidrug Therapies for HIV: Optimal and STI Control Approaches. Mathematical Biosciences and Engineering, 2004, 1(2): 223-241. doi: 10.3934/mbe.2004.1.223
    [8] Hakan Özcan, Bülent Gürsel Emiroğlu, Hakan Sabuncuoğlu, Selçuk Özdoğan, Ahmet Soyer, Tahsin Saygı . A comparative study for glioma classification using deep convolutional neural networks. Mathematical Biosciences and Engineering, 2021, 18(2): 1550-1572. doi: 10.3934/mbe.2021080
    [9] Donggu Lee, Sunju Oh, Sean Lawler, Yangjin Kim . Bistable dynamics of TAN-NK cells in tumor growth and control of radiotherapy-induced neutropenia in lung cancer treatment. Mathematical Biosciences and Engineering, 2025, 22(4): 744-809. doi: 10.3934/mbe.2025028
    [10] Baba Issa Camara, Houda Mokrani, Evans K. Afenya . Mathematical modeling of glioma therapy using oncolytic viruses. Mathematical Biosciences and Engineering, 2013, 10(3): 565-578. doi: 10.3934/mbe.2013.10.565
  • Virotherapy, using herpes simplex virus, represents a promising therapy of glioma. But the innate immune response, which includes TNF-α produced by macrophages, reduces the effectiveness of the treatment. Hence treatment with TNF-α inhibitor may increase the effectiveness of the virotherapy. In the present paper we develop a mathematical model that includes continuous infusion of the virus in combination with TNF-α inhibitor. We study the efficacy of the treatment under different combinations of the two drugs for different scenarios of the burst size of newly formed virus emerging from dying infected cancer cells. The model may serve as a first step toward developing an optimal strategy for the treatment of glioma by the combination of TNF-α inhibitor and oncolytic virus injection.


    1. Introduction

    Oncolytic viruses are genetically altered replication-competent viruses which infect and reproduce in cancer cells but do not harm normal cells. When an infected cell dies many newly formed viruses are released and spread out, infecting neighbouring tumor cells. Treatment by oncolytic viruses (OV) has been and continues to be actively tested in clinical trials for various types of cancer with the use of variety of viruses, [4], [10], [12], [13], [14], including ONYX-15 [7], [11], herpes simplex virus (HSV) [16] and prostate-specific adenovirus CN706 and CN708 [18].

    This therapy, although based on quite promising assumptions, encounters one major obstacle: the innate immune system recognizes the infected cells and destroys them before the viruses within them get a chance to multiply [3].

    It was reported in [6] that CD 163+ macrophages, in rat experiments for glioma, inhibited OV therapy making it unsuccessful. The solution suggested in [6] was to use cyclophosphamide (CPA) as a suppressant of the immune response through the inhibition of CD 163+ and thus enhance the effectiveness of the OV therapy. This approach has been studied from the mathematical point of view by Friedman et al. in [5]. The authors formulated a mathematical model of virotherapy following the earlier work by Wu et al.[21,22], but focusing on the data from glioma rather than from head and neck cancer. The model in [5] was described by a system of PDEs and the effects of the therapy with and without CPA was analysed.

    In the present paper we intend to pick up on this work, but pursue a different avenue based on a very recent paper by Auffinger et al. [2]. In that paper it was suggested that in order to enable the effective action of the virotherapy one should try to block the main "weapon" used by macrophages, namely the TNF-α. It was demonstrated there that inhibition of TNF-α could significantly enhance virus replication and the efficacy of the overall treatment.

    Thus our goal here is to construct a model which captures the interactions between healthy tumor cells, infected tumor cells, viruses, and macrophages and the TNF-α they produce. The model is based on the work of Friedman et al. [5] and the PDE model presented there. However, here we will formulate a reduction of this model from the spatial PDE model to a population type ODE model. This reduction will enable us to pursue a detailed dynamical system analysis of the model as well as analysis of it as an optimal control problem. Although the spatial element has to be compromised for that, it is not entirely removed from the analysis. Indeed, we will be able to estimate the tumor radius in terms of the cells population. The efficacy of both treatments by injection of virus and TNF-α inhibitor will be analysed in the context of the radius of the tumor bringing the spatial aspect back into the picture.

    The approach pursued in our paper will be to target the tumor by combining the two therapies: the viral injection and the TNF-α inhibitor. We will analyze the response of the system to various doses, particularly the efficacy of the therapy, having as a goal the minimization of the tumor radius. The administration of the virus will be pursued through a continuous injection of a fixed dose. This is motivated by the fact that the therapeutic efficacy of systemic chemotherapeutic agents is significantly limited due to the presence of highly selective brain-blood-barriers (BBB). In order to overcome this limitation, researchers have developed a new delivery method, Convention Enhanced Delivery (CED), which relies on intracranial delivery of viral vectors through continuous infusion via catheters. CEDhas recently reached phase Ⅰ/Ⅱ in clinical trials [2], although so far this procedure has been shown to be only moderately effective in recurrent glioma patients [2].

    Both therapeutic agents, the virus and the TNF-α inhibitor, have negative side-effects. Side-effects of virus such as HSV used in tumor virotherapy are significant in rodent tumor models and it is not known if these will occur also in humans [20]. In the case of TNF-α inhibitors the side effects include opportunistic infections like tuberculosis, listerosis and pneumocystis. There are also adverse events including systemic lupus, congestive heart failure and general infections [1]. Because of these side effects, the amount of dosing should be optimized to maximize the effect and minimize the negative side effects. In this paper we will pursue study of the efficacy which is a natural first step towards analysis of this model as an optimal control problem.


    2. A mathematical model for the treatment of glioma with virotherapy and TNF-α inhibitors

    Let x(t) denote the density of cancer cells (uninfected), y(t) the density of infected cancer cells, v(t) the density of the virus, M(t) the density of the macrophages, and T(t) the concentration of TNF-α. The model includes two controls, u1 and u2. The control u1(t) represents the amount of the virus that is injected into the tumor, and the control u2(t) stands for the dosage of the TNF-α inhibitor. The "in" and "out" flows between the compartments and the actions of the two drugs are schematically illustrated in Figure 1.

    Figure 1. Flow chart of the model for virotherapy.

    The burst number is the number of virus that emerge from a dying cancer cell. We shall take it in the range of 50b150. Using unit of gcm3, the parameter b is the burst size defined as b×mass of virusmass of cells=b×106.

    The dynamics of the model is expressed mathematically by the following system of ODEs:

    dxdt=αxβxvδxx, (1)
    dydt=βxvkyTK+Tδyy, (2)
    dMdt=A+syMδMM, (3)
    dTdt=λ1+u2MκyTK+TδTT, (4)
    dvdt=b1kyTK+T+bδyyρxvδvv+u1, (5)

    All the densities and concentrations are in unit of gcm3. In Eq. (1) α represents the proliferation rate of uninfected cancer cells and δx is the death rate; β is the infection rate of x by viruses v. In Eq. (2) the term kyTK+T represents the necrotic death of infected cells caused by TNF-α, while δy is the death by apoptosis. When a cell y dies by apoptosis, b virus particles are released, while if it dies by necrosis a very small number, b1, of viruses emerge. These are accounted by the first two terms on the right hand side of Eq. (5). In Eq. (3) the terms A and δMM represent the source and death of macrophages under healthy normal conditions, while syM accounts for the tumorigenic response of the immune system invoked by the infected cells y. In Eq. (4), the first term on the right-hand side is the production of TNF-α by macrophages, while the remaining two terms are loss by absorption within y cells and loss by natural degradation. The virus equation (5) includes virus particles from dead y cells and loss from absorption by x (i.e. ρxv) and natural degradation/clearance (δvv). We also included in the model a continuous injection u1 of virus, as virotherapeutic drug, and a continuous injection u2 as TNF-α inhibitor (in Eq. (4).

    In the present paper, we take u1const=C and u2const=D, but in future work we shall use u1 and u2 as time-dependent adaptive control variables in order to determine optimal strategies for glioma treatment. Since all the density and concentration in (1)-(5) are in the same unit of gcm3, ρβ=virusmasscellmass is a small number, and similarly κk is a small number. We expect b1 to be very small (viruses are damaged during necrosis) so for simplicity we shall take b1=0. As in [5] the viruses burst (or replication) number b will play a major role in the progression of the disease and its treatment.

    We denote by n(t) the density of all the dead cells. Then, in addition to the dynamics given by Eqs. (1)-(5), we have the equation

    dndt=kyTK+T+δyy+δxx+δMMμn, (6)

    where μ is the rate by which dead cells are cleared out of the tumor.

    Although we do not intend to directly focus on this equation, it will be taken into account in the calculation of the tumor radius.


    2.1. Calculation of the tumor radius

    We assume that the tumor is spherical with variable radius R(t), volume V(t) and mass m(t)=x(t)+y(t)+M(t)+n(t). At each point of the sphere, the density of the cells increases at the rate dmdt. Adding up the equations (1)-(3) and (6) we get that

    dmdt=αxδxx+A+syMμn.

    The total mass of the tumor then increases at rate

    V(t)(α˜xδx˜x+A+s~yMμ˜n)

    where ˜z denotes the average of z. Let θ0 is be the sum of averages, θ0=˜x+˜y+˜M+˜n. We assume that an increase in total mass causes the tumor volume to increase proportionally, that is by θ0dVdt and that ~yM=˜y˜M, so that

    θ0dVdt=V(t)(α˜xδx˜x+A+s˜y˜Mμ˜n).

    Since

    1VdVdt=3RdRdt,

    we get

    θ03RdRdt=(α˜xδx˜x+A+s˜y˜M)μ˜n.

    Hence

    θ03RdRdt=(α˜xδx˜x+A+s˜y˜M)μ(θ0˜x˜y˜M).

    Finally, in order to simplify the calculations, we assume that the averages ˜x, ˜y, ˜M and ˜v satisfy the same equations that x, y, M, v. Hence we get the following formula for the tumor radius:

    θ03RdRdt=(αxδxx+A+syM)μ(θ0xyM). (7)

    2.2. Determination of the parameters

    Table 1 gives the values of the parameters which will be used in our analysis. Explanations concerning the calculations of the parameter values are given in the Appendix.

    Table 1. Parameters of the model.
    ParameterDescriptionNum. valuesDimension
    αProliferation rate of uninfected tumor cells0.21/day
    βInfection rate of tumor cells by viruses2104cm3gday
    ρRate of loss of viruses during infection4102cm3gday
    kEffectiveness of the inhibitory action of TNF-α0.41/day
    δyInfected tumor cell death rate0.21/day
    λTNF-α production rate2.861031/day
    δTTNF -α cell degradation rate55.451/day
    δMMacrophages death rate0.0151/day
    bBurst size of infected cells during apoptosis(50150)106
    b1Burst size of infected cells during necrosis0,b
    KCarrying capacity of the TNF-α5107gcm3
    κDegradation of TNF-α410101/day
    due to its action on infected cells
    δvVirus lysis rate0.51/day
    AConstant source of macrophages9107gcm3day
    sStimulation rate of macrophages by infected cells0.15cm3gday
    without stimulus
    δxdeath rate of uninfected cancer cells0.11/day
    μremoval rate of dead cells0.251/day
    θ0average of total densitiy of cells0.9g/cm3
    Cconstant infusion of the virus(05)107gcm3day
    Dconstant infusion of the TNF-α inhibitor030
     | Show Table
    DownLoad: CSV

    3. Results

    In this section we simulate the model (1)-(7) in order to determine how the state of the system responds to a combined therapy. Values of the parameters for the simulations will be taken from Table 1, unless specified otherwise. We assume that the process starts with the following initial conditions:

    x(0)=0.7,y(0)=0.1,n(0)=0,M(0)=0.1,T(0)=107,4π3R(0)3=0.9,

    so that approximately R(0)=0.6.


    3.1. Behavior of the model without therapy (uncontrolled system)

    We assume that initially a dose of viruses given by v(0)=106 injected and we start by studying the behavior of the system when no additional therapy is given.

    Figure 2 shows the profile of all the variables of the model for the first 50 days for different values of the burst number b. If b is small, namely b=70, the tumor radius increases from its initial value R(0)=0.6 to 43 at T=50, that is, R(50)=43cm. For b=90, R(50) is still large, namely R(50)=6cm. It is only when b=150 that R(50) does no longer increase relative to R(0). These results are in agreement with the mouse experiments in [5] in the sense that if b=50 the tumor radius quickly increases while if b=150 the tumor radius begins to decrease. As in [5], we see also here the oscillations that occur in each of the variables. These oscillations are natural. For example, as y oscillates, so will the viruses which are released from dying infected cells, and so will the rate of macrophages whose activation depends on y.

    Figure 2. Graphs of model variables for C=0,D=0.

    The simulation of T(t) with b=70 suggests that the initial load v(0) results in a massive increase of TNF-α, which later seems to normalize. We also see that after the initial injection of the virus, the virus is multiplying to achieve its maximum by day 28; after that time the virus density keeps decreasing -the drug is 'too weak'. As a result, the growth of R after day 30 is limited. However, the increasing amount of virus in the first 30 days causes M to grow -which leads to growth of T mentioned above. And, as earlier, large amounts of T restrain the replication of virus residing in infected cells y -which is bad for the therapy.

    Moving on to the case with b=90, shown in Figure 2, we observe that the behaviour of T is almost the same as in the previous case. The radius R is still increasing in time, but not as sharply as before. Finally, the case of b=150 presented in Figure 2 shows that the growth of v is faster, but similarly it decays later to finally restart growing again. We can also see that R(t) is first slightly growing, then it decreases, and finally it slightly increases again, mimicking the oscillation of v.

    The observed behavior shows that if we would want to achieve the tumor size reduction without continuous therapy, but just with initial viral injection at time t=0, we would need to inject a very high (probably unrealistic) amount of the virus. This suggests that continuous therapy of virus injection is needed.


    3.2. Behavior of the model with continuous constant viral infusion

    We will now study the behavior of the system assuming that we apply constant viral injection u1(t)C=5107 and the initial load v(0)=C, again for three viral replication numbers, b=70, b=90, b=150. Figure 3 shows the evolution of the system under these assumptions. We see the same oscillating behaviour as before. But, more importantly, we can see the effect of the drug on the tumor radius: for b=70 and b=90 the radius R=R(t) is still increasing, although much less than in the case of C=0 (no continuous infusion). But for the burst number b=150 the tumor radius, after a small initial increase, is strictly decreasing, with R(50) approximately half the initial tumor radius.

    Figure 3. Graphs of model variables for C=5107,D=0.

    In case b=70 the radius R(t) is increasing, but the size after 50 days is almost 5 times smaller then in Figure 2, though we gave smaller initial dose than in the case without therapy. Again v(t) reaches its maximum value (now at day 25), but it does not die out later on. Furthermore, the growth of M, and consequently of T, is smaller, and it has less impact on killing y(t). Heading to the case of b=90, we are still not in the most desirable situation -the radius is still growing, although 3 times less than in Fig. 2 at day 50. As might have been expected, for b=150 the tumor radius is decreasing, but only after a brief increase. This initial increase is probably caused by the fact that the initial virus load is small. The graphs of v(t) suggest that the bigger b is, the earlier the maximum of v is reached, and the larger is its value.

    Figures 4 and 5 show the profile of R(50) as a function of b (for 70b110 in Fig. 4 and 110b150 in Fig. 5) for different values of C (above and below C=5107). We see that R(50) is a monotone decreasing function of b; furthermore, for smaller b's the decline in R(50), as b increases, is more steep. For each value of b, we can determine the exact value of C for which the drug will decrease R(50) below its initial size. For example, if C=5107, then R(50) will be smaller than R(0) only if b>125.

    Figure 4. Graph of R(50) for b[70,110] and different values od C.
    Figure 5. Graph of R(50) for b[110,150] and different values od C.

    Figure 5 suggests that, for b<120, the doses around 107 do not decrease R by day 50 -but for b>120, the smaller doses will make R(50) smaller than R(0) (without using the TNF-α inhibitor D).

    Additionally, in that case we do not have to use D to cure the patient. Similar behavior is shown in Figure 4, for b[70,110]. We can also conclude that the greater C is, the less the effect of additional units of viruses, especially in the second case. Hence, administrating more C and counting that it will solve the problem is not a good idea. R(50) as a function of b, for every fixed C, is of course decreasing, and also it is a convex function, in both cases. However, our goal, which is the tumor size reduction, cannot be fulfilled by applying only constant viral injection, unless we assume very high, probably unrealistic rate of production of the virus.


    3.3. Behavior of the model with combined therapy

    Next we will assume that the combined therapy is given at the constant rates, i.e., u1(t)C and u2(t)D. For our simulations we will take C=5·107 and D=15. Again we will study the response of the system for various levels of viral production by the infected cells. The responses for b=70, b=90, b=150, are shown in Figure 6, which is analogous to Fig. 3.

    Figure 6. Graphs of model variables for C=5107,D=15.

    Firstly, for b=70, the tumor size increases, but not as much as in the case D=0. The graph of T(t) looks similar to the case without D therapy, the values of T(t) are of an order of magnitude lower than earlier -now the maximum is close to 3.5107 in comparison to 7106 where D=0. Applying u2 also supports growth of v -now it is about 5 times higher. What is interesting, is that although T is decreased, M is larger than before.

    For b=90 the tumor radius, after initial increase, goes down to reach its original size. Finally for b=150, we obtain our desired goal for tumor reduction; the radius decreases significantly to about R(50)=0.1cm.

    Now it becomes natural to look closer at the therapy itself and see how the two main agents, viral infusion and TNF-α inhibitor, contribute to the success of the therapy. In other words, what is the contribution of each of them into the goal of reducing the tumor radius. Figure 7 shows how the system evolves in terms of R(50), for variable replication number b, while applying different amounts D. We take C=5107 and study the range of b of [70,90]. We see that even with b=80 we can decrease R(50) from R(0) if we take D=30, that is, if we restrict the production of TNF-α by 97%.

    Figure 7. Graph of R(50) for b[70,90] with different D and fixed C=5107.

    In Figure 8, we take the range of b to be in the interval [110,130] and use a lower C, namely C=3107. Even with such high burst number, with D=1, R(50)>R(0). This case shows the significant effect of small doses of D may have, namely, for b123, with the low dose of D, D=3 (i.e., the production of TNF-α was restricted by only 75%) we get the desired shrinkage of the tumor radius i.e. R(50)<R(0).

    Figure 8. Graph of R(50) for b[110,130] with different D and fixed C=3107.

    Summarizing, our goal of shrinking the tumor size can be achieved under different scenarios, so in choosing the best combination, one should take into consideration potential serious side-effects of the treatment. This brings us to the main topic of this work, which is the evaluation of the efficacy of both treatments.


    3.4. Efficacy of the combined therapy

    In order to capture more clearly the benefits of the dual therapy by C and D, we introduce the concept of efficacy. Let T denote the duration of the therapy. For our analysis we will choose the window of T=50. We denote by R(C,D) the radius of the tumor at the day 50 under the combined treatment with u1(t)C and u2(t)D. The efficacy of the combined therapy is defined by

    E(C,D)=R(0,0)R(C,D)R(0,0)

    Figure 9 is an efficacy map for the case b=90, showing the efficacy of the combined treatment with C varying along the horizontal axis and D along the vertical one. We can see that the efficacy of the combined treatment increases with either C or D. For small C, the efficacy goes up sharply with D. For D>4 (TNF-α production is inhibited by more that 80%) the efficacy increases slowly with C. An efficacy map for b=70 has the same features as in Fig. 9 (not shown here). Thus, for small doses of D the efficacy goes sharply up with C, but for larger doses of D the efficacy increases slowly with C. This confirms the observation made in section 3.3.

    Figure 9. Efficacy map for b = 90.

    Figure 10 is the efficacy map for the case b=150. Here we take a smaller range of C, C[107,5107]. The reason is, that for higher values of C so the effect of D is not that visible; R(50) will become very small even if D=0.

    Figure 10. Efficacy map for b = 150.

    4. Discussion and conclusion

    Virotherapy represents a promising treatment of glioma. However the innate immune response, in particular by macrophages, reduces significantly the effectiveness of the treatment, by killing virus infected cancer cells prematurely. Recent experiments [15] show that blocking macrophages-produced TNF-α can enhance virotherapy treatment in glioma. In the present paper we developed a mathematical model which includes both drugs, oncolytic viruses and TNF-α inhibitor. We observed that the burst number b, that is, the number of new viruses released from the dead cancer cell plays a critical role in the model. Without any treatment, if b is too 'small' the tumor radius will continue to grow, but for larger b, say b>150 the tumor radius will decrease. More generally, given any combined therapy (C,D) and the terminal treatment time T, the tumor radius R(t) at time t=T will be smaller then the initial radius R(0) if and only if b exceeds certain threshold number, b(C,D). Furthermore, we developed an efficacy map of treatment that depends on b. We illustrated it for small b=90 and large b=150. For b=90 the efficacy increases more sharply with the increase in C, whereas for b=150 the efficacy is almost independent on C. This means that for smaller burst number the increase in C has a much larger effect on reducing the tumor radius than if the burst number is larger. Generally the maps show the importance of the TNF-α inhibitor application in the combined treatment. Even a small dose applied we can obtain better efficacy than raising the level of viral infusion very significantly. However, side effects of both drugs in combined treatment will require to limit the amount of doses. Under such limitations the efficacy map could be used as a prognosis tool. In this paper, we use constant amounts of each of the drugs but this does not have to be the case, because the patient situation may require to modify these levels as the treatment goes on. Future work should include developing schemes for evolving treatment of glioma with u1=u1(t)=C(t) and u2=u2(t)=D(t) as functions of time keeping as a goal minimizing the tumor and the side effects. This will require formulating the model as an optimal control problem. We plan to pursue analysis using methodologies as in [19] of this model with the goal to compute optimal time varying regimens for this problem.


    5. Appendix: Parameter estimation

    The model in [5] was based on mice experiments. Prorating some of the parameters to humans, we take

    α=0.2/day,δx=0.1/day,μ=0.2/day.

    We assume that the infected cells die faster than x cells and take δy=0.2/day. The diameter of the virus is 0.13μm. The diameter of the cell is 10μm. Hence the mass of 5105 viruses is approximately equal to the mass of one cell. The mass of the cell is 109g. Hence the mass of one virus is 21015g.

    From [5,8] we get that

    71010<β<1.7108inmm3virusday.

    We take

    β=109mm3virusday

    which in units of cm3gday gives

    β=10810321015cm3gday=2104cm3gday.

    The parameter ρ is in units of cm3cellday. Hence

    ρ=β5105=21045105=4102cm3gday.

    The value of K=5107gcm3 is taken from [17] and the value of δT=55.45/day is taken from [9]. We assume that if T=K then the death rate of y by T is twice the death rate δy, which gives k=0.4/day. The ratio κk is the same as the ratio of mass of Tmass of cell, which we take to be 109. Hence

    κ=109k=0.4109/day=41010/day

    In [9] the death rate of macrophages depends of their phenotype: for M1 phenotype the death rate is 0.02/day and for M2 phenotype, it is 0.008/day. Accordingly we take δM=0.015/day. The source of activated macrophages in the healthy brain is constitutively small; we take A=9107gcm3day. The parameter s is difficult to measure; we take s=0.15cm3gday.

    According to [9] macrophages infected by M. tuberculosis produce TNF-α at the rate λ=1.07103/day. We assume that the macrophages activated by the infected tumor cells are capable of producing TNF-α at a larger rate, taking λ=2.86103/day.


    Acknowledgments

    The research of U. Ledzewicz is partially supported by the National Science Foundation under research Grant No. DMS 1311729. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. A. Friedman has been partially supported by the Mathematical Biosciences Institute (MBI) of The Ohio State University.


    [1] [ C. Antoni,J. Braun, Side effects of anti-TNF therapy: Current knowledge, Clin Exp Rheumatol, 22 (2002): 152-157.
    [2] [ B. Auffinger,A.U. Ahmed,M.S. Lesniak, Oncolytic virotherapy for malignant glioma: Translating laboratory insights into clinical practice, Front. Oncol., 3 (2013): 1-32.
    [3] [ E.A. Chiocca, Oncolytic viruses, Nat. Rev. Cancer, 2 (2002): 938-950.
    [4] [ L.K. Csatary,G. Gosztonyi,J. Szeberenyi,Z. Fabian,V. Liszka,B. Bodey,C.M. Csatary, MTH-68/H oncolytic viral treatment in human highgrade gliomas, J. Neurooncol, 67 (2004): 83-93.
    [5] [ A. Friedman,J. Tian,G. Fulci,E. Chioca,J. Wang, Glioma virotherapy: Effects of innate immune suppression and increased viral replication capacity, Cancer Res., 66 (2006): 2314-2319.
    [6] [ G. Fulci,L. Breymann,D. Gianni,K. Kurozomi,S.S. Rhee,J. Yu,B. Kaur,D.N. Louis,R. Weissleder,M.A. Caligiuri,E.A. Chiocca, Cyclophosphamide enhances glioma virotherapy by inhibiting innate immune responses, PNAS, 103 (2006): 12873-12878.
    [7] [ I. Ganly,D. Kirn,G. Eckhardt,G.I. Rodriguez,D.S. Soutar,R. Otto,A.G. Robertson,O. Park,M.L. Gulley,C. Heise,D.D. Von Hoff,S.B. Kaye,S.G. Eckhardt, A phase I study of ONYX-015, an EiBattenuated adenovirus, administered intratumorally to patients with recurrent head and neck cancer, Clin. Cancer Res., 6 (2000): 798-806.
    [8] [ M.P. Hallsworth,C.P. Soh,S.J. Lane,J.P. Arm,T.H. Lee, Selective enhancement of GM-CSF, TNF-alpha, IL-1 and IL-8 production by monocytes and macrophages of asthmatic subjects, Eur Respir J., 7 (1994): 1096-1102.
    [9] [ W. Hao,E.D. Crouser,A. Friedman, Mathematical model of sarcoidosis, PNAS, 111 (2014): 16065-16070.
    [10] [ K. Jacobsen,L. Russel,B. Kaur,A. Friedman, Effects of CCN1 and macrophage content on glioma virotherapy: A mathematical model, Bull Math Biol, 77 (2015): 984-1012.
    [11] [ F.R. Khuri,J. Nemunaitis,I. Ganly,J. Arseneau,I.F. Tannock,L. Romel,M. Gore,J. Ironside,R.H. MacDougall,C. Heise,B. Randley,A.M. Gillenwater,P. Bruse,S.B. Kaye,W.K. Hong,D.H. Kirn, A controlled trial of ONYX-015, a selectively-replicating adenovirus, in combination with cisplatin and 5-fluorouracil in patients with recurrent head and neck cancer, Nat. Med., 6 (2000): 879-885.
    [12] [ Y. Kim, H. G. Lee, N. Dmitrieva, J. Kim, B. Kaur and A. Friedman, Choindroitinase ABC I-mediated enhancement of oncolytic virus spread and anti tumor efficacy: A mathematical model PLOS ONE 9 (2014), e102499.
    [13] [ R.M. Lorence,A.L. Pecora,P.P. Major,S.J. Hotte,S.A. Laurie,M.S. Roberts,W.S. Groene,M.K. Bamat, Overview of phase I studies of intravenous administration of PV701, an oncolytic virus, Curr. Opin. Mol. Ther., 5 (2003): 618-624.
    [14] [ J.M. Markert, Conditionally replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma: Results of a phase I trial, Gene. Ther., 7 (2000): 867-874.
    [15] [ W.H. Meisen,E.S. Wohleb,A.C. Jaime-Ramirez,C. Bolyard,J.Y. Yoo,L. Russel,J. Hardcastle,S. Dubin,K. Muili,J. Yu,M. Callgiuri,J. Godbout,B. Kaur, The impact of macrophage-and microglia-secreted TNF-α on oncolitic hsv-1 therapy in the glioblastoma tumor microenvironment, Clin Cancer Res., 21 (2015): 3274-3285.
    [16] [ T. Mineta,S. Rabkin,T. Yazaki,W. Hunter,R. Martuza, Attenuated multi-mutated herpes simplex virus-1 for the treatment of malignant gliomas, Nat. Med., 1 (1995): 938-943.
    [17] [ J.C. Oliver,L.A. Bland,C.W. Oettinger,M.J. Arduino,S.K. McAllister,S.M. Aguero,M.S. Favero, Cytokine kinetics in an in vitro whole blood model following an endotoxin challenge, Lymphokine Cytokine Res., 12 (1993): 115-120.
    [18] [ R. Rodriguez,E.R. Schuur,H.Y. Lim,G.A. Henderson,J.W. Simons,D.R. Henderson, Prostate attenuated replication competent daenovirus (ARCA) CN706: A selective cytotoxic for prostate-specific anti-positive prostate cancer cells, Cancer Res., 57 (2000): 2559-2563.
    [19] [ H. Schättler and U. Ledzewicz, Optimal Control for Mathematical Models of Cancer Therapies Springer Publishing Co., New York, USA, 2015.
    [20] [ G. Wollmann,K. Ozduman,A. N. van den Pol, Oncolytic virus therapy for glioblastoma multiforme: Concepts and candidates, Cancer J., 18 (2012): 69-81.
    [21] [ J.T. Wu,H.M. Byrne,D.H. Kirn,L.M. Wein, Modeling and analysis of a virus that replicates selectively in tumor cells, Bull. Math. Biol., 63 (2001): 731-768.
    [22] [ J.T. Wu,D.H. Kirn,L.M. Wein, Analysis of a three-way race between tumor growth, a replication-competent virus and an immune response, Bull. Math. Biol., 66 (2004): 605-625.
  • This article has been cited by:

    1. Elzbieta Ratajczyk, Urszula Ledzewicz, Maciej Leszczyński, Heinz Schättler, Treatment of glioma with virotherapy and TNF-α inhibitors: Analysis as a dynamical system, 2018, 23, 1553-524X, 425, 10.3934/dcdsb.2018029
    2. Talal Alzahrani, Raluca Eftimie, Dumitru Trucu, Multiscale modelling of cancer response to oncolytic viral therapy, 2019, 310, 00255564, 76, 10.1016/j.mbs.2018.12.018
    3. Elzbieta Ratajczyk, Urszula Ledzewicz, Heinz Schättler, Optimal Control for a Mathematical Model of Glioma Treatment with Oncolytic Therapy and TNF-α
    Inhibitors, 2018, 176, 0022-3239, 456, 10.1007/s10957-018-1218-4
    4. Avner Friedman, Xiulan Lai, Francesco Bertolini, Combination therapy for cancer with oncolytic virus and checkpoint inhibitor: A mathematical model, 2018, 13, 1932-6203, e0192449, 10.1371/journal.pone.0192449
    5. A. M. Elaiw, A. D. Al Agha, A reaction–diffusion model for oncolytic M1 virotherapy with distributed delays, 2020, 135, 2190-5444, 10.1140/epjp/s13360-020-00188-z
    6. A.M. Elaiw, A.D. Al Agha, Analysis of a delayed and diffusive oncolytic M1 virotherapy model with immune response, 2020, 55, 14681218, 103116, 10.1016/j.nonrwa.2020.103116
    7. Johannes P. W. Heidbuechel, Daniel Abate-Daga, Christine E. Engeland, Heiko Enderling, 2020, Chapter 21, 978-1-4939-9793-0, 307, 10.1007/978-1-4939-9794-7_21
    8. Handoko Handoko, Setyanto Tri Wahyudi, Ardian Arif Setyawan, Agus Kartono, A dynamical model of combination therapy applied to glioma, 2022, 48, 0092-0606, 439, 10.1007/s10867-022-09618-8
    9. S.N. Antontsev, A.A. Papin, M.A. Tokareva, E.I. Leonova, E.A. Gridushko, Modeling of Tumor Occurrence and Growth-III, 2021, 1561-9451, 71, 10.14258/izvasu(2021)4-11
    10. Salaheldin Omer, Hermane Mambili-Mamboundou, Assessing the impact of immunotherapy on oncolytic virotherapy in the treatment of cancer, 2024, 1598-5865, 10.1007/s12190-024-02139-8
    11. Weizhong Zhang, Zhiyuan Yan, Feng Zhao, Qinggui He, Hongbo Xu, TGF-β Score based on Silico Analysis can Robustly Predict Prognosis and Immunological Characteristics in Lower-grade Glioma: The Evidence from Multicenter Studies, 2024, 19, 15748928, 610, 10.2174/1574892819666230915143632
    12. Dayong Qi, Xueyan Tao, Jiashan Zheng, Boundedness of the solution to a higher-dimensional triply haptotactic cross-diffusion system modeling oncolytic virotherapy, 2025, 25, 1424-3199, 10.1007/s00028-024-01040-y
  • Reader Comments
  • © 2017 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(4280) PDF downloads(586) Cited by(12)

Article outline

Figures and Tables

Figures(10)  /  Tables(1)

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog