Content uploaded by Sulyman Iyanda
Author content
All content in this area was uploaded by Sulyman Iyanda on Dec 10, 2020
Content may be subject to copyright.
@MalariaMath
1192 The spread of drug resistance in P. vivax vs. P. falciparum malaria –
the effect of hypnozoites
Sulyman Iyanda, Kristan A. Schneider
Hochschule Mittweida, University of Applied Sciences
Abstract
Background: Despite the efforts to control and eradicate malaria worldwide, it remains one of the major chal-
lenges to global development. The use of anti-malaria drugs still is the keystone in malaria control and prevention.
However, the spread of drug-resistance in P. falciparum is a serious threat. While many drugs became ineffective
to treat P. falciparum malaria, drug-resistance is uncommon in P. vivax malaria. It has been argued that differen-
cies in the life-histories of the two species, deftermining fitness-components lead to a more effective mechanism
selecting for resistance in P. falciparum . These include in particular the onset of gametocytogenesis and the
longevity of gametocytes. The presence of hypnozoites in P. vivax is a further life-history differfence that has not
been taken into account yet.
Methods: A population-genetic model for the evolutioinary dynamics of anti-malaria drug resistance, that explic-
itly incorporates the sleeping sporozoites in the liver, hypnozoites which make it more general and applicable to
every woman pathogenic malaria species.
Results: By implemeting the model with simulation, it displays how the presence of dormant hypnozoites delays
the evolutionary process underlying drug-resistance in P. vivax compared with P. falciparum. Moreover, the
presence of hypnozoites (and hence relapses) does not affect the fitness driving drug resistance, but only delays
the evolutionary dynamics.
Conclusions: There is need to take into consideration, species specific life history for an effective malaria control
and eradictaion programes. Specificaly, to nullify the challengies being facing when dealing with p.vivax malaria.
The model per se is applicable to all malaria species.
1. Introduction
What triggers the slower spread of Drug resistance in P. vivax compared to P. falciparum?
•Drug resistance evolutionary process;
•Primaquine, Hypnozoites Drug restricted by G6PD defficiency;
•Drugs with resistance:
◦Chloroquine remains choice to combat P. vivax;
◦Sulfadoxine-pyrimethamine;
◦Artemisinin increase concern;
◦Mefloquine.
•Different life history:
◦Onset of gametocytogenesis;
◦Longevity of gametocytes;
◦Dormant parasites, Hypnozoites, in the liver.
•Hypothesis: Hypnozoite, a further life history is responsible for slower spread in P.vivax ;
•Transmission cycle:
◦Similar for all plasmodium species Different life-stage mophology of parasites;
◦2-non-overlaping stages: Asexual/Sexual.
2. Assumptions
•Simple special case model for illustration;
•4-haplotypes formed by 2 bi-allelic loci, a selective and neutral;
•A selective locus with two alleles Asand Arfor sensitive and resistant, respectively;
•Initial frequency of Ar, p0=1
pop.size ,pt→frequency of Arin generation t;
•Host heterogeneity host falling into particular discrete strata;
•One parasite haplotype drop to the host by individual vector;
•Fitness is time independent.
3. Evolutionary Dynamic
Fitness: Probability of sporozoites having gametocyte
offsprings, picked up by mosquito:
•wsand wrare average fitness of sensitive and
resistant alleles, As, Arrespectively;
•Fitness Parameters:
◦c→Metabolic cost;
◦ds→Effect of drug on fitness of As;
◦dr→Effect of drug on fitness of Ar.
•Naturally: ds> dr;
•wr/wsin p. falciparum is greater than wr/wsin
p.vivax;
•s→Selection coefficient, 1 + s=wr
ws
;
•Selection is vary in each class Treated/untreated
•Fitness →Determined by allele at selected locus;
Χ
Χ
Χ
Χ
Χ
Χ
Χ
Χ
haplotype reservoir
mosquito stage human stage mosquito stage
1-α
α
1-α
α
qa relapses from generations t-1, t-2,... q0 infections from generation t
haplotype distributions
from prev. generations
haplotype reservoir
haplotype distributions
from prev. generations
p1(t)
p3(t)
p2(t)
p4(t)
p1(t-a)
p2(t-a)
p3(t-a)
p1(t-a)
p2(t-a)
p3(t-a)
p4(t-a)
p4(t+1)
p3(t+1)
p2(t+1)
p1(t+1)
p4(t-a)
p1(t)
p3(t)
p2(t)
p4(t)
p1(t-a)
p2(t-a)
p3(t-a)
p1(t-a)
p2(t-a)
p3(t-a)
p4(t-a)
p4(t+1)
p3(t+1)
p2(t+1)
p1(t+1)
p4(t-a)
untreated
treated
untreated
treated
Fig1: Model Idealization
•Relapse:
◦Various period of time Weeks, months or even
years;
◦No specific distribution;
◦Absence of vector does not matter.
•Multiplicity of infection (MOI, cf. Fig. 2):
◦Dynamics is independent on MOI;
◦Absence of co-infection but Super-infection.
w(T)
1w(T)
2w(U)
1w(U)
2
(1 −c)(1 −dr) 1 −ds1−c1
Table1: Fitness scale
recomb. generates
new genetic variation
eectively no recomb.
single infection (MOI=1)
x
4 super-infections (MOI=4)
vector-host transmission infection host-vector transmission
& recombination
x
x
Fig2: MOI vs inbreeding
4. Dynamics of P. vivax/P. falciparum
•Evolutionary dynamics depends on wr/wsand
p0=1
pop.size :
◦Earlier gametocytogenesis in p.vivax→initially
more gametocytes from sensitive than resistant
merozoites;
◦Slower spread in p. vivax, due to smaller fitness
ratio wr/ws.
•qa→Proportion of infection relapse from agenera-
tions. in the past (a= 1, . . . , A);
•q0→Probability. of new infection in gen. t;
•Frequency change (cf. fig.1);
pt+1 =
wr
A
X
a=0
qap(t−a)
wr
A
X
a=0
qap(t−a)+ws 1−
A
X
a=0
qap(t−a)!.(1)
•If
A
X
a=0
qa= 1 . . .,vivax dynamics (1) reduce to Falci-
parum (2);
0 2000 4000 6000 8000 10000
0.0 0.2 0.4 0.6 0.8 1.0
generations
frequency of resistance
P. vivax 80% relapses
P. vivax 60% relapses
P. vivax 40% relapses
P. vivax 20% relapses
P. falciparum no relapse
Fig3: Effect of selection coefficient/ relapses
•Frequency change, generation t→t−1:
pt+1 =wrpt
wrpt+ws(1 −pt).(2)
•Numerical examples presented;
•Illustration of effects of relapses (cf. fig.4 & 5);
•A sufficiently small initial frequency considered;
•First resistant parasite occurs in generation 0;
•Dynamics depend on:
◦Relapse proportion;
◦Frequency in present/previous generations;
◦Averagee fitness.
•Generation time (t, discrete) 6=real time;
•Low transmission area larger initial frequency
/more relapses.
0 2000 4000 6000 8000 10000
0.0 0.2 0.4 0.6 0.8 1.0
generations
frequency of resistance
P. vivax 80% relapses
P. vivax 60% relapses
P. vivax 40% relapses
P. vivax 20% relapses
P. falciparum no relapse
Fig4: Effect of initial freq. p0/ relapses
5. Conclusions/Discussion
•Model acounts for relapses→tailored to p. vivax and p. ovale, hence applicable to other plasmodium species;
•Effects of the proportion of relapses from previous generations are clearly visible;
•Initial frequency of mutant alleles, positively correlated to delay of the dynamics but not the case for selection
coeficient;
•Relapses slow-down the spread of resistance as they act as a "seed bank" that brings back the frequency
distribution of the previous generations in the future Hypnozoites;
•Time-average of the past generations responsible for the slow-down of the evolutionary dynamics in P. vivax.
Acknowledgements
This research was supported by grants from the DAAD (“Mathematics against malaria within the AIMS network”, project-ID
57417782), DFG (“Ökologisch nachhaltige Wertschöpfungsketten in in der Landwirtschaft durch Optimierung des Insektizid-
Gebrauchs aufgrund von automatisiertem Schädlings-Monitoring”), and the SMWK (“Vorlaufforschung Technologieentwick-
lung 4.0”).
Reference
[1] K. SCHNEIDER,Charles darwin meets ronald ross - a population-genetic framework for the evolutionary dynamics of
malaria, Submitted, (2019).
[2] SCHNEIDER K. A, ESCALANTE A. A. Fitness components and natural selection: why are there different patterns on the
emergence of drug resistance in Plasmodium falciparum and Plasmodium vivax?, Malaria journal. 2013;12(1):15.
Contact
Sulyman Iyanda
University of Applied Sciences, Mittweida
Technikumplatz 17, 09648, Mittweida, Germany
e-mail: siyanda@hs-mittweida.de
Prof. Dr. Kristan A. Schneider
University of Applied Sciences Mittweida
Technikumplatz 17, 09648, Mittweida, Germany
e-mail: kristan.schneider@hs-mittweida.de
