THE ROLE OF MATHEMATICAL MODELLING IN EPIDEMIOLOGY WITH PARTICULAR REFERENCE TO HIV/AIDS Senelani Dorothy Hove-Musekwa Department of Applied Mathematics NUST- BYO- ZIMBABWE Outline of Talk Aim and objectives Epidemiology Model Building Example Conclusion AIM To bring awareness to medical epidemiologists and pubic health providers of how mathematical models can be used in epidemiology OBJECTIVES: To highlight the purpose of mathematical modelling in epidemiology . To give basic principles on epidemic mathematical modelling To highlight one of the mathematical models which have been developed. Background Empirical Modelling-data driven Application of statistical extrapolation techniques Back calculation method Short term projection only Disadvantages Requires reliable and substantial complete data WHAT IS MATHEMATICAL MODELLING? An activity of translating a real problem into mathematics for subsequent analysis of the real problem Model Development Steps Identify the problem Identify existing knowledge Formulation of Mathematical model No agreement Comparison with The real world (model validation) agreement Report writing Interpretation of solution Mathematical Solution What is epidemiology? DEFINITION:- THE STUDY OF THE DISTRIBUTION, FREQUENCY AND DETERMINANTS OF HEALTH PROBLEMS AND DISEASE IN HUMAN POPULATION PURPOSE:- TO OBTAIN, INTERPRET AND USE HEALTH INFORMATION TO PROMOTE HEALTH AND REDUCE DISEASE BASED ON TWO FUNDAMENTAL ASSUMPTIONS: - HUMAN DISEASE DOES NOT OCCUR AT RANDOM - HUMAN DISEASE HAS CAUSAL AND PREVENTIVE FACTORS THAT CAN BE IDENTIFIED THROUGH SYSTEMATIC INVESTIGATION OF DIFFERENT POPULATIONS OR SUBGROUPS OF INDIVIDUALS WITHIN A POPULATION IN DIFFERENT PLACES OR AT DIFFERENT PLACES OR AT DIFFERENT TIMES KEY QUESTIONS FOR SOLVING HEALTH PROBLEMS WHAT? IS THE HEALTH PROBLEM, DISEASE OR CONDITION, ITS MANIFESTATIONS, CHARATERISTICS WHO? IS AFFECTED:- AGE, SEX SOCIAL STATUS,ETHNIC GROUP WHERE? DOES THE PROBLEM OCCUR IN RELATION TO PLACE OF RESIDENCE, GEOGRAPHICAL DISTRIBUTION AND PLACE OF EXPOSURE QUESTIONS-contd WHEN? DOES IT HAPPEN IN TERMS OF DAYS, MONTHS, SEASON OR YEARS HOW? DOES THE HEALTH PROBLEM DISEASE OR CONDITION OCCUR, SOURCES OF INFECTION, SUSCEPTIBLE GROUPS. OTHER CONTRIBUTING FACTORS QUESTIONS-contd WHY? DOES IT OCCUR IN TERMS OF THE REASONS FOR ITS PERSISTENCE OR OCCURANCE SO WHAT? INTERVENTIONS HAVE BEEN IMPLEMETED AS A RESULT OF THE INFORMATION GAINED, THEIR EFFECTIVENESS, ANY IMPROVEMENTS IN HEALTH STATUS The General Dynamic Of An Epidemic Individuals pass from one class to another with the passage of time. Mathematical model tries to capture this flow by using compartments The purpose of mathematical modelling in epidemiology To develop understanding of the interplay between the variables that determine the course of the infection within an individual and the variables that control the pattern of infection within the communities of people. To provide understanding of the pathophsiology of a disease e.g. HIV. To estimate the incidence and prevalence of a disease e.g.HIV infection in both current and in the past. To identify the groups of the population that are currently at highest risk of contracting a particular disease e.g. HIV. Functions of mathematical models – understanding Explicit assumptions – testable predictions Framework for data analysis Projections Interventions: Outcome Impact Perverse outcomes Combining Interventions Target Setting Impact of new technologies Advocacy Model Example: A TWO-STRAIN HIV-1 MATHEMATICAL MODEL TO ASSESS THE EFFECTS OF CHEMOTHERAPY ON DISEASE PARAMETERS Developed by Shiri, Garira and Musekwa 2005 D epar t m ent of A pplied M at hem at ics N at ional U niv er sit y of Science and T echnology VARIABLE INFECTIOUSNESS OVER THE HIV INFECTION PERIOD Source HIV Insite, University of California San Francisco, School of Medicine http://hivinsite.ucsf.edu D epar t m ent of A pplied M at hem at ics N at ional U niv er sit y of Science and T echnology MODEL ASSUMPTIONS Cell mediated response and no humoral immune response Infection is by two viral strains An uninfected cell once infected remains infected for life Only CD4+ T cells are infected and upon infection cells become productive Treatment drugs: RTIs and Pis act only on the wildtype strain with drug efficacy of RTI and PI respectively Mutant strain viral particles not susceptible to the drug’s antiviral effects No pharmacological and intracellular delays when drugs are administered Eight interacting species Mass action principle employed, i.e., rate at which T cells are infected is proportional to the product of abundances of T cells and viral load • Constant mutation in viral genes would lead to continuous production of viral variants able to evade to some extent the CTL defenses operating. Genetic mutations lead to changes in the structure of viral peptides, i.e. epitopes and these can become invisible to the body’s defenses. If there are too many variants, the immune system becomes incapable of controlling the virus. • Constant mutation in viral genes would lead to continuous production of viral variants able to evade to some extent the CTL defenses operating. Genetic mutations lead to changes in the structure of viral peptides, i.e. epitopes and these can become invisible to the body’s defenses. If there are too many variants, the immune system becomes incapable of controlling the virus. D epar t m ent of A pplied M at hem at ics N at ional U niv er sit y of Science and T echnology STABILITY ANALYSIS Need to remark that the model is reasonable in the sense that no population grows negative and no population grows unbounded The model predicts that within the nonnegative orthant, the number of densities of the seven species attain two steady state values, one with no virus, an uninfected steady state and another with a virus, an endemically steady state Basic reproductive ratio (R0) - the number of newly infected cells that arise from any one cell when almost all cells are uninfected. D epar t m ent of A pplied M at hem at ics N at ional U niv er sit y of Science and T echnology THE BASIC REPRODUCTIVE RATIO The Ratio determines: •Whether an infection can occur determines whether disease will progress or not •Growth rate of infection speed of disease progression •Asymptomatic Period determines time to progress to disease •Necessary effort to control controlling the ratio parameters, we can control the disease T(0) T(1) T(2) R0 = 2 Transmission No Transmission Infectious Susceptible T(0) T(1) T(2) R0 = 1.5 Transmission No Transmission Infectious Susceptible T(0) T(1) T(2) R0 = 2 Transmission No Transmission Infectious Susceptible Immune D epar t m ent of A pplied M at hem at ics N at ional U niv er sit y of Science and T echnology The wild-type strain reproductive ratio is given by The mutant strain’s reproductive ratio is given by D D epar epartt m ment ent of A pplied M at hem at ics N at ional U niv er sit y of Science and and T T echnology echnology CTL EFFECTS 1. CTLs only kill infected cells (a2 = b2 = 0 and h2 ≠ 0), ratio is given by R021 = sα2β2N2 μ2μT(α2 + h2C2 ) 2 . CTLs reduce infection rate of T cells and viral burst size (h2 = 0, a2 ≠ 0 and b2 ≠ 0) -b2C2 R022 = sN2e -a2C2 β2e μ2μT 3 . CTLs kill infected cells and reduce viral infectivity (b2 = 0, a2 ≠ 0 and h2 ≠ 0) -a C sα2N2β2e 2 2 R023 = μ2μT(α2 + h2C2 ) D epar t m ent of A pplied M at hem at ics ics Nat ational ional U Univ niver ersit sityy of of Science Science and and TTechnology echnology N Continued – CTL Effects 4. CTLs kill infected cells and reduce viral burst size (a2 = 0, b2 ≠ 0 and h2 ≠ 0) R024 = sα2N2e -b2C2 β2 μ2μT(α2 + h2C2 ) Comparing the reproductive ratios For α2>>h2C2 and if c2≈∞ •R022 < R021, •R022 <R023 R023 < R021 and and R024 < R021 . R022 < R024 •The hierarchy of the reproductive ratios for a virus with a high rate of viral induced cell killing (high cytopathicity) relative to infected cell CTL mediated killing (α2>>h2C2) and a2<b2 is: R02 < R022 < R024 < R023 < R021 D epar t m ent of A pplied M at hem at ics N at ional U niv er sit y of Science and T echnology The hierarchy of the reproductive ratios for a virus with a low rate of viral induced killing (low cytopathic effect) relative to CTL mediated killing (α2<<h2C2) and a2>b2 is: R02 < R023 < R024 < R021 < R022 Results •Non- lytic effects are critical in the control of virus if the virus’s cytopathic effect is high. •Lytic effects of CTLs are critical in controlling viral load if the virus is less virulent D epar epartt m ment ent of of A A pplied pplied M M at at hem hemat at ics ics D N at at ional ional U U niv niver er sit sit yy of of Science Science and T echnology N NUMERICAL RESULTS A R0-no immune response B C R021-h2≠0 E R022-a2,b2≠0 R023-a2,h2≠ 0 D R024-b2, h2≠0 R02F R02 < R022 < R024 < R023 < R021 < R0 D epar t m ent of A pplied M at hem at ics N at ional U niv er sit y of Science and T echnology ENLARGEMENT OF R1 ASYMPTOMATIC PHASE AIDS CHRONIC PHASE AIDS The parameter values are s2=20,μT=0.02, r2=0.01, β2=0.005,k2=0.0025, BT=350, α2=0.25,h2=0.001,N2=1000, a2=0.015, b2=0.05, μ2=2.5,p=0.00001,2=1.3, BV=400 (Kirschner, 1996; Ho et al. 1995; Dixit and Perelson, 2004; Joshi, 2002) D epar t m ent of A pplied M at hem at ics N at ional U niv er sit y of Science and T echnology Discussion •no chemokines to inhibit infection, no cytokines to reduce burst size with or without killing, there is no clinical latency. •Presence of HIV-1 suppressive factors produced by CTLs control the viral load during HIV infection – thus the presence of chronic phase. •Non-lytic CTL effects are crucial to control viruses with high cytopathicity effects. •Killing of virally infected cells is critical for low cytopathicity viruses. • Results provide evidence to shift our focus to immune based therapies if we are to control the debilitating effects of HIV. • Therapeutic strategies would prompt the body’s own immune system to respond and control HIV infection – immune based therapies should include cytokine modulators and active immunotherapeutics that enhance production of effective cytokines and chemokines by HIV specific CTLs. • Due to the continual generation of new HIV variants that escape CTL killing and resist current ARVs, these therapies should interfere more effectively with the replication and budding processes of the virus. • In conclusion, immune based therapy is the only hope if we are to fight the epidemic. • Constant mutation in viral genes would lead to continuous production of viral variants able to evade to some extent the CTL defenses operating. Genetic mutations lead to changes in the structure of viral peptides, i.e. epitopes and these can become invisible to the body’s defenses. If there are too many variants, the immune system becomes incapable of controlling the virus. •The battle between HIV and the body’s defensive forces is a clash between two armies. Each member of the HIV ARMY is a GENERALIST (able to attack ANY enemy cell it encounters) but each member of the IMMUNE ARMY is a SPECIALIST (able to recognize an HIV SOLDIER if the soldier is waving a flag of a PRECISE COLOUR) WHO IS GOING TO WIN THE WAR? Recommendations Introduce taught courses on the transmission dynamics of diseases, employing some mathematical content in the training of medical doctors and others associated with public health Meanwhile corroboration between health workers, statisticians and mathematicians should be intensified THANK YOU: Partnerships for Health Research &Development Questions? D epar t m ent of A pplied M at hem at ics N at ional U niv er sit y of Science and T echnology Robustness of Model Results Proposition: Any function f(C,ai) where C is a variable (CTL count) and ai is a parameter (i=1,2), where a1 is the effectiveness of each CTL in reducing viral infection of naïve CD4+ T cells and a2 is the effectiveness of each CTL in reducing viral burst size, with the following properties: 1. lim f(C,ai) = 0, C≈∞ 2. lim f(C,ai) = 1, C≈0 3. f(C,ai) is strictly decreasing function of C and 4. f(C,ai) is positive definite can be used to model the effectiveness on non-lytic effects of CTLs in system (1) and still gives the same hierarchy of reproductive ratios, e.g., 1 1+aiC