Characterization of Onchocerca Volvulus

University of Ghana http://ugspace.ug.edu.gh

CHARACTERIZATION OF ONCHOCERCA VOLVULUS RESPONSE TO IVERMECTIN TREATMENT AND IDENTIFICATION OF SINGLE NUCLEOTIDE POLYMORPHISMS ASSOCIATED WITH THESE RESPONSES IN SOME ONCHOCERCIASIS ENDEMIC REGIONS IN GHANA

KWADWO KYEREME FREMPONG (10070961)

THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON

IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE AWARD

OF DOCTOR OF PHILOSOPHY DEGREE IN ZOOLOGY

DECEMBER 2016 University of Ghana http://ugspace.ug.edu.gh

DECLARATION

I do hereby declare that the experimental work described in this thesis was carried out by me except for the references cited from other research works and that this work has not been presented either in whole or in part for any other degree in any institution elsewhere.

KWADWO KYEREME FREMPONG (STUDENT)

DR. MIKE Y. OSEI-ATWENEBOANA (SUPERVISOR)

PROF. MARÍA-GLORIA BASÁÑEZ (SUPERVISOR)

PROF. DANIEL A. BOAKYE (SUPERVISOR)

PROF. EBENEZER O. OWUSU (SUPERVISOR)

University of Ghana http://ugspace.ug.edu.gh

DEDICATION

To the Almighty God, my wife Naa Adjeley Frempong and my son Kwabena Adom O.

Frempong.

University of Ghana http://ugspace.ug.edu.gh

ACKNOWLEGEMENT

I thank the Almighty God for making it possible to successfully complete this dissertation. I would never have been able to complete this dissertation without the guidance of my principal supervisor Dr. Mike Yaw Osei-Atweneboana (Council for Scientific and Industrial Research [CSIR]-Water Research Institute, Accra) and co-supervisors: Prof. Maria Gloria Basanez (Imperial College, London, UK), Prof. Daniel A. Boakye (APOC/Noguchi Memorial Institute for Medical Research, NMIMR) and Prof. Ebenezer O. Owusu (Department of Animal Biology and Conservation Science [DABCS], University of Ghana). I am also grateful to Dr. Martin Walker (Imperial College, London, UK), Prof. Robert A. Cheke (Natural Resources Institute, UK) and Prof. Michael D. Wilson (NMIMR) who were also part of the supervisory team that gave me so much support.

I would like to express my deepest gratitude to the onchocerciasis research team at the CSIR- WRI who gave me support both on the field and in the laboratory: Edward Jenner Tettevi, Francis Balunna Veriegh, Samuel Armoo, Bright Idun, Rafik Mohammed, Nana Asor, Isaac Frimpong as well as staff and students of the Bio-Medical and Public Health research unit at CSIR-WRI, Accra. To staff and students of the Parasitology Department, NMIMR who supported me one way or the other, I am very grateful.

My appreciation also goes to Dr. Glover, Department of Pediatrics, University of Ghana Medical School (UGMS), Korle-Bu, Accra for carrying out the nodulectomies and Dr. Simon Atta, Department of Microbiology, UGMS, Korle-Bu, Accra for his support during the parasitological work in the laboratory. The field works would not have been possible without the support from the various community leaders, ivermectin distributors, volunteers and participants of this study. To staff of the Neglected Tropical Disease Control Programme, Ghana Health Services for providing me with the longitudinal treatment coverage data. I also acknowledge financial support from the Royal Society-Leverhulme Trust Africa Award, The European Initiative for Neglected Tropical diseases and the WHO-CARIRS Project. I appreciate the administrative support from the Head and staff of the DABCS, University of Ghana. Finally to my wife, family and friends for their moral support to complete this thesis.

iii

University of Ghana http://ugspace.ug.edu.gh

TABLE OF CONTENTS

DECLARATION ...... i

DEDICATION ...... ii

ACKNOWLEGEMENT ...... iii

TABLE OF CONTENTS ...... iv

LIST OF FIGURES ...... x

LIST OF TABLES ...... xii

LIST OF APPENDICES...... xiii

LIST OF ABBREVIATIONS...... xiv

ABSTRACT ...... xvi

CHAPTER ONE ...... 1

GENERAL INTRODUCTION ...... 1

1.1 Background ...... 1

1.2 Rationale/Justification for the Study ...... 7

1.3 Hypothesis ...... 9

1.4 Objectives: ...... 10

1.4.1 Main objectives ...... 10

1.4.2 Specific objectives ...... 10

University of Ghana http://ugspace.ug.edu.gh

CHAPTER TWO ...... 11

LITERATURE REVIEW ...... 11

2.1 Epidemiology of Onchocerciasis ...... 11

2.1.1 Distribution of onchocerciasis infections ...... 11

2.1.2 Risk factors of onchocerciasis ...... 15

2.1.3 Socio-economic importance of onchocerciasis ...... 15

2.1.4 Burden of onchocerciasis ...... 16

2.1.5 Transmission of onchocerciasis ...... 17

2.2 Clinical Symptoms of Onchocerciasis ...... 19

2.2.1 Skin disease manifestations ...... 19

2.2.2 Blindness...... 22

2.2.3 Hanging groin ...... 22

2.2.4 Defect of the Central Nervous System (Nodding & Nakalanga Syndromes) .... 23

2.3 Diagnostic Methods and Epidemiological Indices of Onchocerciasis ...... 24

2.3.1 Onchocerciasis diagnostic methods ...... 24

2.3.2 Epidemiological indices for onchocerciasis assessment ...... 25

2.4 Control of Onchocerciasis ...... 27

2.4.1 History of onchocerciasis control ...... 27

2.4.2 Chemotherapy in the control of onchocerciasis ...... 30

2.4.3 Ivermectin (macrocyclic lactone) mechanism of action ...... 33

University of Ghana http://ugspace.ug.edu.gh

2.4.4 Responses of Onchocerca volvulus to ivermectin ...... 35

2.5 Mathematical Modelling of Onhcoerciasis Infection ...... 37

2.6 Onchocerciasis Vector Species in Ghana...... 39

2.7 The Human Onchocerciasis Parasite (Onchocerca volvulus) ...... 41

2.8 Candidate Proteins for Ivermectin Resistance ...... 43

2.8.1 P-glycoprotein structure and function ...... 43

2.8.2 Beta-tubulin structure and function ...... 45

CHAPTER THREE ...... 46

GENERAL METHODOLOGY...... 46

3.1 Ethical Approval ...... 46

3.2 Study Sites...... 46

3.3 Study Design and Parasitological Methods ...... 49

3.4 Isolation of Worms...... 51

3.5 Molecular Studies ...... 51

CHAPTER FOUR ...... 54

RESPONSES IN ONCHOCERCIASIS PATIENTS AFTER BIANNUAL IVERMECTIN

TREATMENT IN GHANA ...... 54

4.1 Introduction ...... 54

4.2 Methods ...... 56 vi

University of Ghana http://ugspace.ug.edu.gh

4.2.1 Data analyses ...... 56

4.3 Results ...... 60

4.3.1 Trends in community infection ...... 61

4.3.2 Trends in microfilarial repopulation ...... 64

4.3.3 Microfilarial repopulation rates ...... 64

4.4 Discussion ...... 68

CHAPTER FIVE ...... 73

EMBRYOSTATIC EFFECT OF IVERMECTIN ON ONCHOCERCA VOLVULUS ADULT

FEMALE WORM ...... 73

5.1 Introduction ...... 73

5.2 Methods ...... 76

5.2.1 Study sites and design ...... 76

5.2.2 Nodulectomy ...... 77

5.2.3 Nodule digestion, isolation and examination of adult worms ...... 78

5.2.4 Embryogram procedure ...... 79

5.2.5 Data Analyses ...... 80

5.3 Results ...... 81

5.3.1 Regression analysis of annual treatment on normal microfilariae production ... 91

5.4 Discussion ...... 94

vii

University of Ghana http://ugspace.ug.edu.gh

CHAPTER SIX ...... 101

SINGLE NUCLEOTIDE POLYMORPHISMS WITHIN BETA-TUBULIN AND P-

GLYCOPROTEIN GENES ASSOCIATED WITH ONCHOCERCA VOLVULUS SUB-

OPTIMAL RESPONSES TO IVERMECTIN TREATMENT ...... 101

6.1 Introduction ...... 101

6.2 Methods ...... 105

6.2.1 Characterization of adult female worm responses (selection of worms) ...... 105

6.2.2 DNA extraction from adult female worms ...... 106

6.2.3 Estimation of DNA concentration ...... 108

6.2.4 DNA amplification ...... 108

6.2.5 Gel electrophoresis ...... 111

6.2.6 Data analysis ...... 111

6.3 Results ...... 112

6.4 Discussion ...... 122

CHAPTER SEVEN ...... 129

GENERAL DISCUSSION, CONCLUSION AND RECOMMENDATIONS ...... 129

7.1 General Discussion ...... 129

7.1.1 Strengths and limitations of the study ...... 135

7.2 General Conclusion ...... 136

7.3 Recommendations ...... 137 viii

University of Ghana http://ugspace.ug.edu.gh

REFERENCES ...... 138

APPENDICES ...... 153

University of Ghana http://ugspace.ug.edu.gh

LIST OF FIGURES

Figure 2.1: Countries covered by onchocerciasis control (and elimination) programmes…...13

Figure 2.2: Global distribution and status of preventive chemotherapy for onchocerciasis in 2015…………………………………………………………...14

Figure 2.3: Life cycle of Onchocerca volvulus……………………………………………….17

Figure 2.4: Palpable nodule…………………………………………………………………..20

Figure 2.5: Immuno-pathological responses in onchocerciasis patients due to presence and death of microfilariae in the skin……………………………………………..21

Figure 2.6: Map of West Africa showing onchocerciasis prevalence in survey villages before and after the implementation of the Onchocerciasis Control Programme in West Africa (OCP)...... 29

Figure 2.7: P-glycoprotein conformational changes during drug efflux……………………...44

Figure 3.1: Map of study the sites in Brong-Ahafo and Northern Regions of Ghana………..48

Figure 3.2: Schematic timeline and illustrative history of participants (Study design)………52

Figure 4.1: Trends in community microfilarial prevalence (CMFP) in the 10 communities after a biannual ivermectin treatment…………….………………..62

Figure 4.2: Trends in community microfilarial loads (CMFLs) in the 10 communities after a biannual ivermectin treatment…………………………………………….63

Figure 4.3: Trends in mean numbers of microfilariae per participant in 9 communities after a biannual ivermectin treatment…………………...……………………….65

Figure 4.4: Six-month microfilarial repopulation rates in 9 communities after a biannual ivermectin treatment……………………………………………………66

Figure 4.5: Relative 6-month microfilarial repopulation rates in 9 communities over the first 2 rounds of a biannual ivermectin treatment…………..………………..67

Figure 5.1: Nodules excised from an onchocerciasis patient…………………………………83

Figure 5.2: Adult female and male worms……………………………………………………83

Figure 5.3: Adult male worm (O. volvulus) showing the coiled head region………………...84

Figure 5.4: Different forms of stretched microfilariae found in the uteri of adult female worm……………………………………………..…………...... 85 x

University of Ghana http://ugspace.ug.edu.gh

Figure 5.5: Embryonic forms of microfilariae found in the uterus of an adult female worm...86

Figure 5.6: Percentage distribution of adult female worms with their reproductive status in the 10 communities……………………………….………………...... 89

Figure 5.7: Mean number of degenerate and normal intra-uterine microfilariae per female worm in the 10 study communities………………………...………...89

Figure 5.8: Frequency distribution of the reproductive status of female worms by their age…………………………………………………………………………..90

Figure 5.9: Percentage distribution of the reproductive status of female worms by their age…………………………………………………………………………..90

Figure 5.10: Relationship between the odds ratio of obtaining normal mf and years of annual treatment in the 10 communities………………………...…………...93

Figure 6.1: Gel electrophoresis pictures showing DNA fragments……………………...... 113

Figure 6.2: Combined genotype frequencies within the three ivermectin responding groups at beta-tubulin positions 1308 and 1545………………..………………121

University of Ghana http://ugspace.ug.edu.gh

LIST OF TABLES

Table 3.1: Longitudinal cohorts of participants in 10 Ghanaian communities who were followed up and skin snipped over the first two rounds of biannual treatment with ivermectin………………………...…………………..53

Table 4.1: Key features of the Log-Linear Marginal Regression Models used to describe the observed microfilarial counts in the longitudinal cohort…..……..59

Table 4.2: Prevalence and intensity of onchocerciasis infection at baseline and end of Study……………………………………………………………………………....60

Table 5.1: Summary results on nodules and worms obtained from participants……………..82

Table 5.2: Summary results on the reproductive status of female worms examined through embryogram…………………..…………………………………………..88

Table 5.3: Summary output of generalized linear mixed model (effect of community on normal microfilariae production)…………...……………………..92

Table 6.1: Primer sequences and expected diagnostic band sizes for amplification………..110

Table 6.2: Genotype frequencies at five SNP positions for beta-tubulin……………………115

Table 6.3: Genotype frequencies at five SNP positions within P-glycoprotein gene……….116

Table 6.4: Genotype frequencies within Kintampo/Pru and Kpandai Districts for three SNP positions…………………………………………………………..118

Table 6.5: Estimated values of inbreeding coefficient (F-statistic) for Kintampo/Pru and Kpandai Districts…………………………………………………………….119

Table 6.6: Hardy-Weinberg Equilibrium at 10 SNP positions identified within beta-tubulin and p-glycoprotein genes……………...……………………………120

xii

University of Ghana http://ugspace.ug.edu.gh

LIST OF APPENDICES

Appendix I: Consent form and ethical clearance certificates………………………………..153

Appendix II: Estimation of community microfilarial loads, prevalence and confidence intervals; Marginal regression model………………………...…..161

Appendix III: Community Microfilarial Load of Onchocerca volvulus at Different Time Points…………………………………………………………………..166

Appendix IV: Community Microfilarial Prevalence of Onchocerca volvulus at Different Time Points ……………………………………………………………….....167

Appendix V: Six-month Onchocerca volvulus microfilarial skin repopulation rates at Different Time Points………………..……………………………….168

Appendix VI: Chromatograms of DNA sequence showing SNP positions…………………169

Appendix VII: Estimation of genotype frequencies and F-statistics………………………..180

Appendix VIII: Publication (Clinical Infectious Disease Journal)………………………….183

xiii

University of Ghana http://ugspace.ug.edu.gh

LIST OF ABBREVIATIONS

ABC: Adenosine triphosphate-binding cassette

APOC: African Programme for Onchocerciasis Control

ATP: Adenosine triphosphate

β-tub: Beta-tubulin

CDC: Centers for Disease Control, Atlanta, USA

CDTI: Community-directed treatment with Ivermectin

CIDs: Community Ivermectin Distributors

CMFL: Community microfilarial load

CMFP: Community microfilarial prevalence

DALYs: Disability-adjusted life-years

DEC: Diethylcarbamazine

DNA: Deoxyribonucleic acid

ELISA: Enzyme-Linked Immuno-Sorbent Assay

EPIONCHO: Onchocerciasis transmission model (Population-based)

GABA: Gamma aminobutyric acid

GHS: Ghana Health Service

GluCl: Glutamate-gated chloride channels

GTP: Guanosine-5'-triphosphate

HWE: Hardy-Weinberg Equation

ICT: Immuno-chromatographic test

IgG: Immunoglobulin G

ITD: Iodide Transport defect

LBV: Lower Black Volta xiv

University of Ghana http://ugspace.ug.edu.gh

MDA: Mass drug Administration

NBD: Nucleotide Binding Domain

NTD: Neglected Tropical Disease

NTDCP: Neglected Tropical Disease Control Programme

OCP: Onchocerciasis Control Programme

OEPA: Onchocerciasis Elimination Program for the Americas

ONCHOSIM: Onchocerciasis micro-simulation model (Individual-based)

PCR: Polymerase Chain Reaction

P-gp: P-glycoprotein

RDT: Rapid Diagnostic Test

SNP: Single Nucleotide Polymorphism

TM: Trans-membrane

TMD: Trans-membrane domain

UV: Ultraviolet

WA1: Walker 1

WHO: World Health Organization

University of Ghana http://ugspace.ug.edu.gh

ABSTRACT

In Ghana, onchocerciasis control with mass ivermectin treatment began in 1987. Despite over two decades of interventions, the disease remains persistent with reports of sub-optimal/poor parasite responses to the drug. Some treated patients are observed with higher microfilarial repopulation rates in skin than expected, an indication of sub-optimal response. Although ivermectin is still effective in reducing microfilaridermias, it is uncertain if its embryostatic effect has been compromised. This thesis was to assess the impact of the first 3 years of biannual treatment strategy in Ghana and quantify responses to standard dose of ivermectin in hosts’ skin, assess the drug’s effect on the reproductive capacities of adult female worms and explore any genetic changes in beta-tubulin (β-tub) and P-glycoprotein (P-gp) genes that are believed to be associated with poor response phenotype. The study was carried out in 10 sentinel communities which had received between 15 and 24 years of annual treatment. A community-wide skin snipping was performed on 956 consenting adults aged ≥20 years to assess the community prevalence and intensity of microfilariae. A cohort of 217 participants who were microfilaria positive and/or had palpable nodules at baseline were followed up over the first two rounds of biannual treatment to estimate the rates of microfilarial repopulation.

Nodulectomies were performed on consenting participants three months after the third round of treatment. Adult worms (male and female) were isolated from nodules using the collagenase technique. Embryogram analyses were performed and adult female worms classified into three response groups (good, intermediate and poor). DNA was extracted from

60 worms which accurately fitted the response classifications. Polymerase Chain Reaction

(PCR) amplifications were performed using specific primers for one region within β-tub and 6 regions within P-gp genes. The amplified products were sequenced and analysed for single nucleotide polymorphisms (SNPs) associated with these responses. The biannual treatment xvi

University of Ghana http://ugspace.ug.edu.gh

substantially reduced infection intensities in most communities, although infections were detected in all communities even after 4 or 5 rounds of biannual treatments. Asubende,

Kyingakrom and New-Longoro communities were identified (all having been previously recognized as responding sub-optimally to ivermectin) with statistically significantly high microfilarial repopulation rates. A total of 225 nodules were excised from 106 participants with an average of about 2 female worms and 1 male worm per nodule. A significantly higher number of female worms (72%) were observed without normal/viable microfilariae

(P < 0.0001). There were no clear associations between the years of annual ivermectin treatment prior to biannual and microfilarial repopulation rates or reproductive status of female worms. A multiple sequence alignment showed 10 SNPs that were polymorphic and analysed for any associations. Three of these SNPs were statistically significantly associated with a poor response phenotype i.e. two at positions 1308C/T (P = 0.016) and 1545A/G (P =

0.008) within β-tub and one at position 5546A/G (P = 0.023) within P-gp. Within the β-tub, there was selection at position 1308C/T and some genotypes were present in good responders but absent in poor responders, vice versa. The heterozygosity was found to be reduced within worms sampled from Kintampo/Pru districts compared to those from Kpandai district.

Although the biannual treatment in Ghana has made an impact, transmission still exists within some communities, and this is suggested to be driven by a few sub-optimally responding female worms in each community. These sub-optimal responses are also associated with some level of genetic changes. Regular monitoring of parasite responses to ivermectin treatment is necessary to avoid a completely resistant population emerging. Based on these findings, it is uncertain if increasing the frequency of ivermectin treatment (in Ghana) will be sufficient to meet the World Health Organization’s goals of onchocerciasis elimination by 2025.

xvii

University of Ghana http://ugspace.ug.edu.gh

CHAPTER ONE

GENERAL INTRODUCTION

1.1 Background

Human onchocerciasis, commonly known as river blindness, is a parasitic disease caused by infection with the filarial worm Onchocerca volvulus Leuckart (Bradley et al., 2005). The parasite is transmitted to humans by infectious blackflies of the genus Simulium Latreille

(Crosskey, 1967), which ingest the first stage offspring of the parasite (microfilariae) from humans, and allow their development to the third stage larvae (L3, infective stage). In the human body, the larvae moult from L3 to L4 soon after inoculation and form a new nodule or join a pre-existing one (Duerr et al., 2001) in the subcutaneous tissue or elsewhere in the body. The larvae develop to L5 (immature adults) and finally mature into adult worms which become reproductively active. After mating, the female adult worm can release up to 1,000 microfilariae per day (Engelbrecht and Schulz-Key, 1984). The microfilariae migrate through the nodules to the epidermis of the skin and likely the skin lymphatics where they are ingested by a blackfly during a blood meal to continue the cycle. Occasionally, at high infection intensity, the microfilariae may be found in the peripheral blood, urine and some body fluids

(CDC, 2013a). Although the disease is rarely considered as life-threatening (Little et al.,

2004a; Walker et al., 2012), it causes chronic ocular and dermal morbidity and severe disability, including blindness. It is the presence of the microfilariae that results in the clinical manifestations associated with onchocerciasis, such as skin disease and itching of the skin, impaired vision, blindness, etc. due to the body’s immune response against the parasite (Ali et al., 2003) and/or its Wolbachia pipientis endobacteria (Brattig, 2004).

University of Ghana http://ugspace.ug.edu.gh

Onchocerciasis is the world's second-leading infectious cause of blindness after trachoma

(WHO, 1995). The vast majority of infections occur in sub-Saharan Africa (99% of cases), with the remainder in some foci in Central and South America, and in the Yemen (WHO,

2008). It has been reported that about 37 million people are thought to be infected, with 90 million at risk in Africa and more than 400,000 infected in Central and South America

(Basáñez et al., 2006).

In Ghana, onchocerciasis is endemic in 9 out of its 10 regions, affecting 3,204 communities within 74 districts. Most of these communities are in the Brong-Ahafo, Ashanti and Northern regions. It is estimated that a population of 3.2 million people are at risk of the disease (Taylor et al., 2009). An epidemiological study carried out in 2004-2005 within 20 endemic communities in Ghana showed an average microfilarial prevalence of 19% (Osei-

Atweneboana et al., 2007). This information indicates that there are residual infections and on-going transmission in some endemic communities in Ghana, though there have been many years of mass treatment with ivermectin and, in the savannah zones, up to 26 years of vector control by the Onchocerciasis Control Programme (OCP) in West Africa.

Onchocerciasis control strategies have evolved significantly since 1974, when control started under the umbrella of the OCP until 2002, when the programme closed officially (Amazigo and Boatin, 2006; Hotez, 2007). The OCP employed weekly larviciding of vector breeding sites in river rapids. Some of the larvicides used were Temephos (Abate®) and Chlorphoxim which were organophosphates as well as a biological control agent, Bacillus thuringiensis serotype H-14, a spore-forming bacterium (Leveque, 1989). This strategy of controlling the larvae was effective in interrupting transmission, but deemed not feasible to be employed in

University of Ghana http://ugspace.ug.edu.gh

the whole of Africa. Therefore, the OCP focused on the savannah areas, where the prevalence of blindness was greater (Hougard et al., 2001).

The introduction and registration of ivermectin (Mectizan®) in 1987 for the treatment of human onchocerciasis, and the decision by Merck and Co. to donate the drug for onchocerciasis control for as long as necessary, represented a great breakthrough for the control of onchocerciasis (Samba, 1994; Thylefors, 2008). Since then, ivermectin has been used for the control of onchocerciasis primarily using an annual strategy within endemic regions by a Community-Directed Treatment with Ivermectin (CDTI) approach and still remains the drug of choice for safe mass treatment that can effectively eliminate the microfilarial stages of O. volvulus (Taylor et al., 1989; Duke et al., 1992; Diawara et al.,

2009). In recent times, insecticides have been used on a much smaller geographical scale to control the disease focally (e.g. the eradication of the Bioko form of S. yahense on Bioko, and the elimination of S. neavei in some foci in Uganda) [Garms et al., 2009; Traore et al., 2009].

Ivermectin acts on nematode parasites (and some ectoparasites) by binding selectively with high affinity to the glutamate-gated chloride ion channels in their muscle and nerve cells

(Wolstenholme and Rogers, 2005). This causes an increase in the permeability of the cell membrane to chloride ions and results in hyper-polarization of the cell, leading to paralysis and death of the parasite (Blackhall et al., 1998; Wolstenholme and Rogers, 2005). The mechanism works very well against the microfilariae of O. volvulus and eventually leads to their death (microfilaricidal effect). However, ivermectin does not kill the adult worms in the same fashion. Rather, the drug temporarilly prevents the intra-uterine release of microfilariae

University of Ghana http://ugspace.ug.edu.gh

from the adult female worm for a number of months (embryostatic effect) [Basáñez et al.,

2008].

There have been reports of sub-optimal responsiveness of the O. volvulus adult female worm to ivermectin, presumably due to continuous usage of the drug (Awadzi et al., 2004a; Awadzi et al., 2004b; Osei-Atweneboana et al., 2007). However, although widespread in nematode parasites of veterinary importance (Kaplan, 2004), ivermectin resistance has not been confirmed in O. volvulus. These reports have been based on the persistence of microfilariae in skin of patients some few months after treatment. Since the standard dose of ivermectin (150-

200 µg/kg body weight) is still capable of clearing >99% of skin microfilariae one month after treatment, the persistence of microfilariae in the skin after treatment has been attributed to a reduced embryostatic effect on some adult female worms in the population (Awadzi et al., 2004a; Awadzi et al., 2004b; Osei-Atweneboana et al., 2007; Basáñez et al., 2008;

Churcher et al., 2009). These female worms recover from the effect of treatment and repopulate host skin with new microfilariae at faster rates (3-4 months) than usually expected after treatment (9-12 months). These adaptations have been reported to have some genetic basis contributing to the reduced sensitivity of some O. volvulus populations to ivermectin treatment (Bourguinat et al., 2007; Prichard, 2007; Osei-Atweneboana et al., 2012).

Due to these reports on sub-optimal responses in Ghana, the Neglected Tropical Disease

Control Programme (NTDCP) of the Ghana Health Services (GHS) assessed the status of onchocerciasis in Ghana in 2009 and in 2010 adopted a biannual (twice a year) ivermectin treatment strategy in some (44 out of 74) endemic communities to address the persistent

University of Ghana http://ugspace.ug.edu.gh

microfilaridermias and faster skin repopulation rate or poor response in some of O. volvulus populations (Turner et al., 2013a; unpublished GHS-NTD report in 2010).

The possibility of resistance of O. volvulus to ivermectin has been of much concern to control programmes since this is the main drug that is currently being used to control the disease on a large scale. In some parasitic nematodes, a number of genes have been reported to be associated with poor response to ivermectin treatment. These include: Permeability- glycoprotein (P-gp), Glutamate-gated chloride channel, P-gp-like proteins, beta-tubulin (β- tub) [isotype 1] and Gamma aminobutyric acid (GABA). Among these genes, P-gp and β-tub have been reported to be strongly associated with poor response of O. volvulus to the drug

(Eng and Prichard, 2005). P-gp 1 is a multidrug resistance protein 1 or ATP-binding cassette which belongs to the sub-family B member 1 (ABCB1). This protein is used by resistant parasites to pump harmful drugs out of their cells and eventually reduce concentration of the drug in the cells (Ughachukwu and Unekwe, 2012). This mechanism has been observed in O. volvulus populations responding sub-optimally to ivermectin treatment (Ardelli et al., 2005;

Eng and Prichard, 2005). On the other hand tubulin is one of several members of a small family of globular proteins and beta-tubulin forms part of the cytoskeletal structure of the adult worm which is present in the cytoplasm of the cell. β-tub has also been associated with sub-optimal responses by O. volvulus populations to ivermectin treatment (Eng and Prichard,

2005; Bourguinat et al., 2007; Osei-Atweneboana et al., 2012).

Ghana hopes to eliminate onchocerciasis at the country level by 2025 and reach a stage where

MDA will be terminated, and control programmes will emphasize the phase of monitoring/surveillance as well as treat individual cases in clinical or community settings

University of Ghana http://ugspace.ug.edu.gh

(Taylor et al., 2009). The World Health Organization (WHO), in its Neglected Tropical

Diseases (NTD) Roadmap of 2012, has also set the goal of elimination of onchocerciasis in selected African countries by 2020 (WHO, 2012a) and in majority (about 80%) by 2025

(WHO, 2012b). These goals can only be achieved if detailed epidemiological research is carried out to track progress of the control programmes in the countries. To this end, it is important for onchocerciasis control programmes to use decision-support transmission models to estimate elimination thresholds before terminating MDA. This is because if MDA is terminated prematurely, there could be recrudescence of the infection and disease. In addition, in those foci where sub-optimal ivermectin responses are confirmed, control programmes may need to deploy alternative treatment strategies to reach elimination. On the other hand, valuable resources may be wasted if ivermectin distribution continues after the prevalence and transmission intensity of the infection have become lower than the elimination threshold. This threshold is also known as the transmission breakpoint, as shown in a recent comparison of the ONCHOSIM and EPIONCHO transmission mathematical models (Stolk et al., 2015).

The purpose of this thesis is to assess the current epidemiological situation of onchocerciasis in Ghana and to identify some single nucleotide polymorphisms (SNPs) in O. volvulus P-gp and β-tub genes associated with poor responses to ivermectin treatment. It is hoped that the results of this thesis will provide useful information for parameterizing mathematical models that will be used in ascertaining the feasibility of onchocerciasis elimination in Ghana. In addition, the information will be needed in the development of molecular genetics tools for early identification of poor response to ivermectin treatment in O. volvulus populations.

University of Ghana http://ugspace.ug.edu.gh

1.2 Rationale/Justification for the Study

The socio-economic burden of onchocerciasis in Africa in general, and Ghana in particular, is enormous. In Ghana, apart from the Greater Accra Region, all the other nine regions are endemic with onchocerciasis and 3.2 million people are at risk (Taylor et al., 2009). The disease has been earmarked for elimination by the WHO in a number of endemic African countries by 2020 / 2025 (WHO, 2012a; WHO, 2012b). It is expected that there will be a complete change in the control strategy from annual to semi-annual treatment in most communities using community-directed treatment with ivermectin (CDTI), especially in the sub-optimally responding villages. As mentioned above, transmission still persists in some communities (Lamberton et al., 2015) and sub-optimal responses to treatment have been reported in some individuals/populations (Awadzi et al., 2004a; Awadzi et al., 2004b; Osei-

Atweneboana et al., 2007; Osei-Atweneboana et al., 2011) despite many years of ivermectin treatment in endemic regions. The current goals of the control programme in Ghana are to move towards interruption of transmission followed by elimination of the infection reservoir

(elimination of the disease as a public health problem can more easily be achieved with annual CDTI (Turner et al., 2014a).

Declaration of elimination is based on current data on infections within the endemic regions as evidence of elimination. Since the last community-wide epidemiological study took place in 2004/2005 (Osei-Atweneboana et al., 2007), this thesis will provide current and up-to-date information on onchocerciasis infections in some endemic regions in Ghana. Data generated will be useful in providing information on Ghana-specific parameters such as community microfilarial prevalence (CMFP), community microfilarial loads (CMFL), repopulation rates

University of Ghana http://ugspace.ug.edu.gh

and genetic information. These are valuable data that will be needed to calibrate mathematical models to explore the feasibility of onchocerciasis elimination in Ghana or predict elimination timelines. Such models when developed will rely on available and current data for validation.

The overall purpose of this thesis is, therefore, to assess the epidemiological and parasitological profiles of O. volvulus in Ghana in order to determine the current status of onchocerciasis in some endemic communities, with particular reference to those showing sub- optimal responses. The current information will also be helpful to the NTDCP of the Ghana

Health Service (GHS) to ascertain whether Ghana is on target to eliminate the public health significance of onchocerciasis within some endemic communities by 2020 and at the country level by 2025, or if more control efforts are needed to meet this target.

Although there have been reports on sub-optimal/poor response of O. volvulus to ivermectin treatment, most of these studies have identified the responses phenotypically, through long- term follow up of treated patients within endemic communities. Such studies involve regular ivermectin treatments and measuring of parasitological responses through skin snipping

(microfilarial repopulation rates) and/or embryogram (macrofilarial reproductive status). It is therefore important that in addition to these phenotypic measures, efforts are focused on developing molecular diagnostic tools that allow direct detection of good and poor responding worms (from microfilariae in skin snips or adult worms from nodules) without the need for long-term follow ups. As part of this study, single nucleotide polymorphisms (SNPs) within the P-gp and β-tub genes of O. volvulus adult female worm associated with poor responses to ivermectin treatment will be identified. These SNPs in the various endemic communities and the frequencies with which they occur will be analysed. This information is important to help

University of Ghana http://ugspace.ug.edu.gh

understand the rate of development of any sub-optimal responses or decreased ivermectin efficacy in susceptible populations that may be indicative of emerging resistance. These will also help to monitor resistance development in naïve O. volvulus populations and also ascertain how widespread genes associated with poor ivermectin responses are in other parasite populations. The SNP information will also be used in the future to develop a molecular diagnostic tool for early detection of poor responses. These molecular tools have become important to save time and costs in detecting poor responses from parasites in any

O. volvulus population. The availability of such a tool has become even more necessary recently since some endemic countries are progressing towards onchocerciasis elimination and may need to monitor ivermectin responses within the parasite population(s) to prevent or mitigate against recrudescence of the infection.

1.3 Hypothesis

Transmission of onchocerciasis has been on-going in Ghana despite many years of vector control and ivermectin treatment in endemic communities. It is, therefore, hypothesized that sub-optimal responses may occur in treated individuals within such communities and that single nucleotide polymorphisms (SNPs) within the P-glycoprotein and/or beta-tubulin genes of O. volvulus adult female worm may be associated with such poor/sub-optimal responses to ivermectin treatment. This hypothesis will be contrasted with other (programmatic) explanations such as poor therapeutic coverage in the study villages.

University of Ghana http://ugspace.ug.edu.gh

1.4 Objectives:

1.4.1 Main objectives

The main aim of this thesis is to determine the current epidemiological status of onchocerciasis in some endemic communities in Ghana and assess phenotypic responses to ivermectin treatment associated with genetic changes in O. volvulus populations.

1.4.2 Specific objectives

The specific objectives of this study are:

1) To assess the current epidemiological status of onchocerciasis in some endemic

communities in Ghana which have received annual ivermectin treatment for prolonged

periods.

2) To evaluate the impact of the biannual strategy of ivermectin distribution in Ghana on

infection prevalence and intensity at community level.

3) To quantify the rates of reappearance of microfilariae in the skin of human hosts after

the introduction of biannual treatment strategy and investigate whether this can

successfully address concerns of sub-optimal responses.

4) To characterize, by embryogram, the reproductive status (phenotype) of adult female

worms.

5) To determine any association between adult female O. volvulus phenotypes and

genetic polymorphisms in the beta-tubulin and P-glycoprotein genes.

University of Ghana http://ugspace.ug.edu.gh

CHAPTER TWO

LITERATURE REVIEW

2.1 Epidemiology of Onchocerciasis

2.1.1 Distribution of onchocerciasis infections

The World Health Organization (WHO) currently estimates that at least 25 million people are infected with onchocerciasis and 120 million people live in areas that put them at risk of infection. The disease is the second-leading infectious cause of blindness world-wide (WHO,

2015b). About 300,000 people are blind because of the parasite and another 800,000 have visual impairment (WHO, 2015a). More than 99% of infected persons live in 31 countries within sub-Saharan Africa and these include: Angola, Benin, Burkina Faso, Burundi,

Cameroon, Central African Republic, Chad, Republic of Congo, Ivory Coast, Democratic

Republic of the Congo, Equatorial Guinea, Ethiopia, Gabon, Ghana, Guinea, Guinea-Bissau,

Kenya, Liberia, Malawi, Mali, Mozambique, Niger, Nigeria, Rwanda, Senegal, Sierra Leone,

South Sudan, Sudan, Togo, Uganda and the United Republic of Tanzania. The rest of the infected people (<1%) are found in some foci within Latin America and the Yemen (WHO,

2015a).

Transmission of onchocerciasis has now stopped in 11 out of 13 foci within Latin America, but it continues within one focal area in Venezuela and one in Brazil, namely the Amazonian focus straddling Venezuela and Brazil (CDC, 2013b). Currently the disease has been successfully eliminated in a number of foci within Latin America and distributions of ivermectin there have ceased. Transmission of onchocerciasis has now been declared eliminated in Colombia (West et al., 2013), Ecuador (Lovato et al., 2014), Mexico 11

University of Ghana http://ugspace.ug.edu.gh

(Rodriguez-Perez et al., 2015) and Guatemala (Sauerbrey et al., 2018) [Figure 2.1]. In the savannah areas of West Africa, blindness from onchocerciasis infection was very common.

The risks of visual impairment increase, in part, as the prevalence and intensity of microfilarial infection increase in the various communities (Little et al., 2004b). A number of communities were abandoned due to increased prevalence of blindness from onchocerciasis and accompanying vector biting rate (intense vector nuisance). The forest areas tend not to record as much blindness compared to the savannah areas, but the comparison is not clear-cut

(Cheke and Garms, 2013). Even with relatively similar intensities of onchocerciasis infection, onchocercal skin disease tends to predominate in the forest, with lower blindness rates

(Dadzie et al., 1990). The different savannah and forest epidemiological patterns are thought to be due to the existence of two strains of O. volvulus. The high prevalence of ocular disease with the savannah strain has been, in part, associated with higher loads of Wolbachia pipientis

(Tamarozzi et al., 2011), which are endosymbiotic bacteria found within the O. volvulus adult worms and microfilariae. They are known to be essential for the filarial worm's fertility and survival (Higazi et al., 2005). In Africa, transmission of onchocerciasis has been interrupted in a number of foci within some endemic countries (Higazi et al., 2013; Katabarwa et al.,

2014) and elimination achieved within three foci in Mali and Senegal (Traore et al., 2012) and two foci in Nigeria (Tekle et al., 2012).

Nonetheless, a recent estimation using the prevalence of palpable nodules in adult males

(criteria by Rapid Epidemiological Mapping of Onchocerciasis, REMO) identified 18 out of

20 countries under the African Programme for Onchocerciasis Control (APOC countries shown in Figure 2.1) as high risk areas (i.e. prevalence of nodules >20%). These areas ranged from small isolated foci to a vast contiguous endemic area of 2 million km2 and 86 million

University of Ghana http://ugspace.ug.edu.gh

people were estimated to live within these high risk areas (Noma et al., 2014). The current distribution of onchocerciasis worldwide; indicating infected countries at different levels of endemicity is shown in Figure 2.2.

Figure 2.1: Countries covered by onchocerciasis control (and elimination) programmes Onchocerciasis control programmes were established in Africa and Latin America to control infections. The Onchocerciasis Control Programme in West Africa (OCP) operated from 1974-2002 to control infections within 11 West African endemic countries. The African Programme for Onchocerciasis Control (APOC) established in 1995 and closed in December 2015. It helped to monitor control activities within Central and East African endemic countries (19) and some few West African countries still endemic for the infection. The Onchocerciasis Elimination Program for the Americas (OEPA) established in 1991 to step up elimination of onchocerciasis in 13 endemic foci within 6 countries in Latin America. Source: http://www.who.int/mediacentre/factsheets/fs095/en/ Accessed on 12/08/2016

University of Ghana http://ugspace.ug.edu.gh

Figure 2.2: Global distribution and status of preventive chemotherapy for onchocerciasis in 2015 Source: http://www.who.int/mediacentre/factsheets/images/Onchocerciasis_2015.png?ua=1; Accessed on 12/08/2016

University of Ghana http://ugspace.ug.edu.gh

2.1.2 Risk factors of onchocerciasis

The major risk of onchocerciasis infection is living near streams or rivers which serve as breeding sites for the Simulium blackflies (CDC, 2013a). Most of the areas where the blackflies are found are rural agricultural areas in sub-Saharan Africa. Many infective bites are often needed before one can be infected. As a result, people who travel for short periods of time (generally less than 3 months) to endemic areas have a low chance of becoming infected with the O. volvulus parasite (CDC, 2013a). People who also work close to rivers that breed the blackflies are of higher risk of getting the infection than those who live far away from the breeding sites. Usually farmers (involved in rural agriculture) and fishermen are of greater risk of getting the infection (Akogun, 1999).

2.1.3 Socio-economic importance of onchocerciasis

The socio-economic impact of onchocerciasis in endemic communities cannot be overlooked.

Inhabitants have been forced to move from fertile land close to river valleys to the less fertile upland due to the disease and nuisance from bites of blackflies (WHO, 2015a). Many young men migrate to urban areas, reducing the productivity of the community and disrupting family life. Those with onchocercal skin infections are stigmatized and their social involvement in the community is limited. In some cases, it can affect chances of even finding a life partner

(Ubachukwu, 2006).

People with the disease often have low self-esteem and the itching results in self-frustration with lack of concentration at work or school (Wogu and Okaka, 2008). Skin infection and aging of skin also makes infected individuals appear aged and unhappy (Ubachukwu, 2006).

Presence of visible nodules on traders reduces their activities and loss of customers engaged

University of Ghana http://ugspace.ug.edu.gh

in trading. Ocular lesion impairing vision resulting loss of jobs, and reduced economic productivity. (Wagbatsoma and Okojie, 2004; Ubachukwu, 2006).

2.1.4 Burden of onchocerciasis

Onchocerciasis-associated ocular disease (visual impairment and blindness) has been identified as the greatest contributor to the burden of onchocerciasis. The next most important contributors are skin lesions and severe itching caused by the death of microfilariae. Some regions do not experience much blindness, but the skin lesions from onchocerciasis have been found to be a major cause of socio-economic burden in terms of disability-adjusted life-years

(DALYs) (Coffeng et al., 2014a). Current DALY estimations do not take into account onchocerciasis-associated epilepsy, nodding disease or excess mortality.

Due to the duration of the infection (adult females have a mean life span of 12-15 years and produce microfilariae for about 9-11 years (Plaisier et al., 1991)), an infected person would have to go through years of continuous treatment which sometimes results in treatment fatigue. The microfilariae which are the main cause of the major and most severe clinical manifestations of onchocercal disease, live predominantly in the skin but may migrate to various other organs and tissues of the body, especially the eye thereby causing blindness

(Hall and Pearlman, 1999; Marty, 2013). Onchocerciasis also affects the immediate family of infected individuals, the community and the entire country. Dependants of affected individuals may suffer the effect of not receiving support due to blindness or other manifestations that may affect the productivity of the infected person. In some cases, children or adults are unable to complete their education or work effectively since they spend time taking care of the sick or the blind (Ubachukwu, 2006).

University of Ghana http://ugspace.ug.edu.gh

The financial burden of treatment of onchocerciasis e mass distribution of ivermectin through the donation of Mectizan® cannot be overlooked. So far a substantial amount of investment has been made in controlling this disease in Africa (Coffeng et al., 2013).

2.1.5 Transmission of onchocerciasis

Figure 2.3: Life cycle of Onchocerca volvulus Source: http://www.cdc.gov/dpdx/onchocerciasis/ Accessed on 12/08/2016

University of Ghana http://ugspace.ug.edu.gh

The transmission cycle of onchocerciasis in humans begins with an infected blackfly (genus

Simulium) with the O. volvulus infective stage (L3) larvae (Figure 2.3). The blackfly bites during the day and the females are involved in transmission in the process of obtaining a blood meal. They use the proteins and irons in the blood as source of nutrients to develop their ovaries to become matured (Crosskey, 1990). During the blood meal, the infected blackfly introduces the L3 into the skin of the human host. The larvae are inoculated into the host and they migrate to the subcutaneous tissues. In the subcutaneous tissues the L3 moult to fourth-stage larvae (L4) within 3-7 days (Schulz-Key and Soboslay, 1994). Then, after a variable period of some weeks, they moult again to the juvenile adult stage (L5). They then mature into full adults and mate after approximately a year and then start producing microfilariae (Duke, 1980). The stretched microfilariae exit through the vulva into the tissues and begin their migration to the skin and other parts of the body including the eyes. They are later picked from the skin by the blackly through a blood meal to begin the cycle.

University of Ghana http://ugspace.ug.edu.gh

2.2 Clinical Symptoms of Onchocerciasis

The clinical manifestations of onchocerciasis is due to the presence of microfilariae in the skin (immunological responses). The Wolbachia endobacteria that are present in the microfilariae also induce more immune responses when they are released into host after the death of microfilaraie. The clinical manifestation of onchocerciasis is wide ranged and includes skin lesions (infections), ocular lesions, lymphatics (hanging groin), defect of the central nervous system (Nodding & Nakalanga syndromes).

2.2.1 Skin disease manifestations

Adult worms are found in nodules (onchocercomata) which may manifest as palpable

(Figure 2.4) or hidden in deep tissues. Nodules are firm, movable, non-tender and often round.

They are normally found in the subcutaneous tissue of the skin and over bony crevices such as the skull, scapula, ribs (Figure 2.4), elbows, iliac crests, sacrum, or knees. The clinical manifestations of onchocerciasis begin with immunopathological responses to microfilariae in the skin (Ali et al., 2003), which tend to be milder in ‘generalized onchocerciasis’

(characterized by high microfilarial loads) and hyper-reactive in ‘sowda’ (accompanied by low microfilaridermia) (Bradley et al., 2005). The severity of skin disease is, in part, related to the presence of Wolbachia pipientis which is known to aggravate the immune responses within the host by virtue of its release of pro-inflammatory cytokines from dying microfilariae

(Higazi et al., 2005; Tamarozzi et al., 2011), and partly due to the presence of filarial antigens

(Brattig, 2004). Skin lesions (onchodermatitis) result in itching and the formation of papular lesions due to scratching and inflammation. Excessive scratching can lead to bleeding and secondary infections (Figure 2.5A). As parasites interact with the immune system, oedema may be formed and when chronic infections are not treated, the skin may appear dimpled with

University of Ghana http://ugspace.ug.edu.gh

a texture like an orange peel (Marty, 2013), known as ‘orange skin’. The skin eventually losses its elasticity and appears prematurely aged, wrinkled and extremely thin (skin atrophy)

[Figure 2.5B and Figure 2.5C]. A later stage of this manifestation appears like a ‘lizard’s skin’. Other immune manifestations may result in loss of melanin pigment in the skin

(depigmentation of skin) which is often referred to as ‘leopard skin’ (Figure 2.5D).

Onchocerciasis-associated skin lesions have been categorized into a standardized clinical classification and grading system for comparative purposes to include: acute and chronic papular onchodermatitis, lichenified onchodermatitis, skin atrophy, depigmentation (leopard skin), and lymphatic involvement such as hanging groin (Murdoch et al., 1993).

Figure 2.4: Palpable nodule The arrow shows palpable nodule on the side of an onchocerciasis patients. Source: http://www.hxbenefit.com/wp-content/uploads/2012/10/Onchocerciasis-Image.jpg Accessed on 16/08/2016

University of Ghana http://ugspace.ug.edu.gh

A B

C D Figure 2.5: Immuno-pathological responses in onchocerciasis patients due to presence and death of microfilariae in the skin Presence of microfilariae in skin can cause itching. Excessive scratching may lead to bleeding in skin and secondary infections (A). If patients are not treated, the skin eventually losses its elasticity and appear aged, wrinkled and extremely thin (B and C). Other immunological manifestations may result in loss of melanin pigment in the skin (depigmentation of skin) which is often referred to as the leopard skin (D). Source: http://emedicine.medscape.com/article/1109642-overview#a2 Accessed on 16/08/2016

University of Ghana http://ugspace.ug.edu.gh

2.2.2 Blindness

The well-known ocular manifestation of onchocerciasis, blindness, begins with inflammatory responses in the anterior chamber of the eye due to the presence and death of microfilariae in the eye, with the subsequent release of somatic filarial and wolbachial antigens as described above. This results in inflammatory processes around the dead microfilariae that start as

‘punctate keratitis’, which may resolve with ivermectin treatment. The conflagration of numerous of these lesions with time and microfilarial intensity, which tend not to resolve lead to ‘sclerosing keratitis’. This results in scarring or clouding of the cornea (opacification) and eventually leads to blindness (Hall and Pearlman, 1999; Ali, 2006). Uveitis, iridocyclitis and cataract may also ensue. In addition to these ‘anterior segment’ lesions, other lesions of the posterior segment include optic nerve atrophy (Bradley et al., 2005).

2.2.3 Hanging groin

Some rare complications of onchocerciasis include hanging groin and hernia (Nelson, 1958;

Dozie et al., 2007). These conditions are triggered by the death of microfilariae that induces intense inflammatory reactions. Infected individuals develop rashes, severe itching and various skin lesions. The skin therefore wastes away and loses elasticity, causing hanging groin. The hanging groin presents as a sac of atrophic skin containing sclerosed inguinal or femoral lymph glands. This condition predisposes the infected individual to hernia (Nelson,

1958).

University of Ghana http://ugspace.ug.edu.gh

2.2.4 Defect of the Central Nervous System (Nodding & Nakalanga Syndromes)

Onchocerciasis infection has been associated with a neurological condition called Nodding

Syndrome. This disease was first documented in the United Republic of Tanzania in the

1960s, then later in the Republic of South Sudan in the 1990s and in northern Uganda in 2007

(Dowell et al., 2013). The disease affects children between the ages of 5 and 15 years. It causes progressive cognitive dysfunction, neurological deterioration, stunted growth and a characteristic nodding of the head. Nodding Syndrome is still known to occur in the southern region of the United Republic of Tanzania, South Sudan and northern Uganda (Dowell et al.,

2013). The disease was associated with an epileptic syndrome caused by Onchocerca volvulus, since prevalence of both onchocerciasis and epilepsy in the areas affected by

Nodding disease was high (Kaiser et al., 2015). The disease is caused by an autoimmune response to parasitic proteins such as leiomodin-1. Studies have shown that antibodies that bind to leiomodin-1 in humans also attach to proteins from Onchocerca volvulus (Friedrich,

2017; Johnson et al., 2017a; Johnson et al., 2017b). Leiomodin-1 protein found in human brain cells have been shown to be similar in structure to that of O. volvulus. Therefore antibodies developed against the parasite after infection also attack the brain cells which lead to the Nodding disease (Friedrich, 2017; Johnson et al., 2017b). A similar condition to the

Nodding disease is known as “Nakalanga Syndrome” which was described in Uganda.

Nakalanga Syndrome is also characterized by growth retardation, physical deformities, endocrine dysfunction, mental impairment and epilepsy in addition to the head nodding

(Foger et al., 2017).

University of Ghana http://ugspace.ug.edu.gh

2.3 Diagnostic Methods and Epidemiological Indices of Onchocerciasis

2.3.1 Onchocerciasis diagnostic methods

Onchocerciasis cases are mainly diagnosed parasitologically by detection, identification and microscopic enumeration of microfilariae in skin snips. Skin snips (2 or more) are often taken from iliac crests, back, calf, etc. using a (2 mm Holth or Walser corneoscleral) punch and placed in a small volume of saline or other medium in a well of a microtitration plate. The microfilariae are counted after 30 minutes and also 24 hours later using an inverted microscope. The microfilariae in samples could also be fixed (with formaldehyde) and dried

(preserved) for future staining with Mayer’s haemalum and analysis (Kale, 1978).

Microfilaridermia is usually measured as the mean number of microfilariae per skin snip

(mf/ss) or, if the snips are weighed, microfilariae per mg of skin (mf/mg). Skin snip examination is not sufficiently sensitive for detection of early infections (in which a substantial population of microfilariae has not yet built in the skin) or for accurate diagnosis in persons with low microfilarial densities. This makes the test a poor choice for assessing the success of control programmes especially at the stage nearing elimination, when levels of microfilaridermia will be very low. The procedure is also inconvenient and invasive.

Other detection procedures include the skin patch test which involves the topical application of diethylcarbamazine (DEC) cream and examination of the skin after 48 hours (Stingl et al.,

1984). A localized papular reaction (known as the Mazzotti reaction) observed may indicate the presence of microfilariae. Although this method is noninvasive, it is somewhat inconvenient for use in the field as it relies on delayed hypersensitivity reactions (and possibly loss to follow up), cannot detect early infections (for the reasons given above), and it does not provide an accurate readout of the intensity of microfilarial infection in the skin.

University of Ghana http://ugspace.ug.edu.gh

Some immunological assays such as the immuno-chromatographic test (ICT) or the Enzyme- linked immuno-sorbent assay (ELISA) using the Ov16 recombinant antigen to detect antibodies (IgG4) in serum of onchocerciasis patients have been used (Weil et al., 2000;

Lipner et al., 2006). Currently, PATH diagnostic technology has developed a rapid diagnostic test (RDT) using Ov16 which is commercially available and is undergoing extensive evaluation in the field in comparison with Ov16 IgG4 ELISA, skin snips, and skin snip PCR.

Infection can be detected within 3 months of infection and more than a year before microfilariae detection in the skin is possible by skin snipping, leading to the conclusion that the Ov16-based test may be able to detect pre-patent infection (Lobos et al., 1991). However,

Ov16 is not a parasite stage-specific antigen and may also be located in adult worms.

Polymerase chain reaction (PCR) based procedures using DNA from skin snips and specific primers have also been used (Fink et al., 2011). These methods are very sensitive and specific but expensive and mainly used for research, although the new WHO (2016) guidelines for stopping ivermectin MDA (WHO, 2016) recommend their use in decision algorithms to inform cessation of treatment and commencement of surveillance.

2.3.2 Epidemiological indices for onchocerciasis assessment

Epidemiological indices are utilized to assess the prevalence and intensity of infection within affected communities. Microfilarial prevalence (the proportion of examined individuals with skin microfilariae) and microfilarial density (the mean number of microfilariae per skin snip or per mg of skin in those examined) in a representative sample of the population in endemic communities (all age groups and typically those above 5 years of age) are two important epidemiological indices. Two additional indices (measured in adults) are the prevalence of palpable nodules (onchocercomata, where the mature worms live) in a sample of adult males

University of Ghana http://ugspace.ug.edu.gh

(the basis of REMO) [Noma et al., 2002], and the community microfilarial load (CMFL), the geometric mean number of microfilariae per skin snip in those (positive and negative) individuals aged 20 years and older (Remme et al., 1986).

University of Ghana http://ugspace.ug.edu.gh

2.4 Control of Onchocerciasis

2.4.1 History of onchocerciasis control

Onchocerciasis control strategies have evolved significantly over the years since 1974, when control started under the umbrella of the Onchocerciasis Control Programme in West Africa

(OCP) until 2002, when the programme closed officially (Amazigo and Boatin, 2006; Hotez,

2007). It covered initially 7 and finally 11 countries in West Africa (Benin, Burkina Faso,

Ivory Coast, Ghana, Guinea, Guinea-Bissau, Mali, Niger, Senegal, Sierra Leone and Togo).

The OCP started as a vector control programme only, applying weekly larviciding to vector breeding sites in river rapids within the savannah areas where blindness rates were high

(Hougard et al., 2001). In 1987 ivermectin was licensed for human use and the first community trials were conducted in Asubende, Ghana. Since 1987, ivermectin has been used to complement vector control or as the only control measure for onchocerciasis such as in the

Western Extension of the OCP, where annual and biannual strategies were compared in foci of Mali and Senegal.

Before (the implementation of) the OCP, the prevalence of onchocerciasis within many of the endemic countries ranged from hypoendemic (<30%) to highly hyperendemic (≥80%), with a recently published Bayesian geostatistical mapping exercise showing the distribution of endemicities in the former OCP area at its commencement (O'Hanlon et al., 2016).

Figure 2.6A shows the location of survey villages within the OCP. These villages were chosen because of their high initial endemicity and to serve as sentinel sites for the monitoring of the progress of the programme. When the OCP ended in 2002, it had succeeded in eliminating onchocerciasis as a public health problem in 10 out of the 11 countries where it operated (Figure 2.6B). The exception was Sierra Leone due to years of armed conflict. Other

University of Ghana http://ugspace.ug.edu.gh

countries like Benin, Ghana, Guinea and Togo still have ongoing transmission (WHO, 2009).

However, Togo is making great progress towards elimination. On the other hand, on-going transmission is still being reported in Senegal, areas of Mali, and in Ivory Coast, the latter probably due to recrudescence of infection because of internal (WHO, 2009).

University of Ghana http://ugspace.ug.edu.gh

B Figure 2.6: Map of West Africa showing onchocerciasis prevalence in survey villages before and after the implementation of the onchocerciasis. Control Programme in West Africa (OCP) A: Before OCP, the prevalence within the majority of these endemic villages was ≥60% because sampling protocols were biased in favour of highly endemic villages and because of monitoring purposes. B: When the OCP ended in 2002, it had succeeded in eliminating onchocerciasis as a public health problem in 10 out of the 11 countries where it operated, with the prevalence in survey villages declining considerably, except in Sierra Leone due to years of armed conflict. Source: http://www.who.int/apoc/onchocerciasis/ocp/en/ Accessed on 12/08/2016

University of Ghana http://ugspace.ug.edu.gh

The African Programme for Onchocerciasis Control (APOC) was launched in 1995 to cover all the remaining onchocerciasis endemic areas in Africa (Amazigo, 2008) and the

Onchocerciasis Elimination Program for the Americas (OEPA) was launched in 1991

(Sauerbrey, 2008) to oversee the elimination of onchocerciasis in the endemic regions of

Latin America (Figure 2.1). Sierra Leone, Ghana, Guinea-Bissau and Ivory Coast were added to the operations of APOC to strengthen onchocerciasis surveillance systems within these countries. APOC closed officially in December 2015 and OEPA is still on-going to support elimination efforts in the Amazonian focus straddling Venezuela and Brazil (Botto et al.,

2016).

2.4.2 Chemotherapy in the control of onchocerciasis

Ivermectin still remains the drug of choice with effective microfilaricidal effect (Taylor et al.,

1989a; Duke et al., 1992; Diawara et al., 2009). In Africa, it is given as an annual in some endemic areas or biannual dose of 150-200 µg/kg body weight. At this dose it markedly helps to control and prevent morbidity (Turner et al., 2014a), inhibits adult worm reproduction for some months, and reduces skin microfilarial loads, maintaining them at low levels, hence reducing net transmission to vectors (Duke et al., 1992; Alley et al., 1994; Osei-Atweneboana et al., 2007; Basáñez et al., 2008). However, the suppression of microfilariae in the skin during the inter-treatment period is not sufficient to completely halt transmission (Turner et al., 2014b). Treatment with repeated standard doses (quarterly and possibly 6-minthly) or at a higher than the 150-200 µg/kg standard dose markedly impairs female worm fecundity and reduces the life span of adult worms (Gardon et al., 2002; Cupp et al., 2004). Treatment is implemented through community-directed treatment with ivermectin (CDTI). Since 2006 in

University of Ghana http://ugspace.ug.edu.gh

Ghana onchocerciasis control has been implemented in the context of the Neglected Tropical

Disease Control Programme (NTDCP) [Taylor et al., 2009].

Although ivermectin is the drug used for mass treatment, there are other microfilaricides that have been used for treatments in the past but have limitations as some of them may incur adverse effects. One of these is diethylcarbamazine (DEC), which causes severe side effects in patients with high microfilarial loads (Haarbrink et al., 1999) and can exacerbate ocular lesions because it leads to a rapid death of microfilariae in the skin and the eyes. Some macrofilaricides (compounds capable of killing the adult worms) have also been used to treat onchocerciasis. Though effective, some of them also have serious side effects. An example is suramin, which in addition requires intra-venous administration (Thylefors and Rolland,

1979). The drug amocarzine has also recorded some successes, although its macrofilaricidal properties are not optimal (Awadzi et al., 1997).

Currently, the tetracycline antibiotic doxycycline is under consideration for the treatment of onchocerciasis patients in clinical settings or in the community following test-and-treat protocols. This antibiotic works by killing the Wolbachia endobacteria which have been found to be in symbiotic association with O. volvulus. By killing the bacteria, the infra-population of adult worms experiences a decline in fertility (they become unable to produce microfilariae because of the sterilizing effects of doxycycline). Additionally, there is a slow decline in survival (thereby avoiding the severe adverse effects associated with a fast killing). This eventually would reduce transmission (Hoerauf et al., 2008; Turner et al., 2010).

Additionally, although not microfilaricidal, skin microfilariae depleted of Wolbachia because of doxycycline treatment are unable to develop successfully in the blackfly vectors (Albers et

University of Ghana http://ugspace.ug.edu.gh

al., 2012). Drawbacks of treatment with doxycycline are the necessary prolonged treatment courses (4–6 weeks at a daily dose of 100–200 mg) that are required to deplete the Wolbachia bacteria low enough to exert its sterilizing and macrofilaricidal effects. However, community trials of treatment with doxycycline have been conducted successfully in Cameroon with high levels of coverage and adherence to treatment (Wanji et al., 2009), and their impact on community levels of infection have been evaluated (Tamarozzi et al., 2012). It is conceived that doxycycline can be used on a test-and-treat basis to tackle onchocerciasis in loiasis co- endemic areas (where MDA with ivermectin cannot be safely delivered to individuals with high loiasis microfilaraemia), in areas with sub-optimal responses to ivermectin and in mop- up elimination settings.

Mathematical modelling has been used to estimate the reduction in adult worm life expectancy from an average of 10 years to 2–3 years (i.e. adult worm lifespan is reduced by

70 – 80%) (Walker et al., 2015). Other possibilities once licensed for human use is the distribution of moxidectin, which has been trialled and in Phase II and III clinical trials, has demonstrated a stronger microfilaricidal effect and a longer suppression of microfilariae in the skin than those of ivermectin (Awadzi et al., 2014). The epidemiological impact and cost- effectiveness of moxidectin have been modelled and results indicate that annual moxidectin could be as effective as biannual ivermectin in reducing timeframes to elimination of onchocerciasis (Turner et al., 2015).

University of Ghana http://ugspace.ug.edu.gh

2.4.3 Ivermectin (macrocyclic lactone) mechanism of action

The macrocyclic lactones (avermectins and milbemycins) are products or chemical derivatives of soil microorganisms belonging to the genus Streptomyces. The avermectins in commercial use are ivermectin, abamectin, doramectin, eprinomectin and selamectin.

Commercially available milbemycins are milbemycin oxime and moxidectin (Vercruysse,

2014). The macrocyclic lactones have a potent, broad antiparasitic spectrum at low dose levels. They are active against many immature nematodes and ectoparasitic arthropods and non ectoparasitic insects such as mosquitoes (Pooda et al., 2015). A single therapeutic dose can persist in concentrations sufficient to be effective against incumbent nematode infections for prolonged periods after treatment (Vercruysse, 2014). The mode of action of ivermectin is similar to the avermectin/milbemycins group. Therefore, the mechanistic action on parasites explained below may apply in most cases to the avermectin/milbemycin group. Ivermectin was initially introduced as a commercial product for animal health in 1981 (Campbell, 2012).

It is effective against a wide range of parasites, including gastrointestinal roundworms, lungworms, mites, lice and hornflies (Burg et al., 1979; Chabala et al., 1980; Campbell et al.,

1983) as well as fleas which cause tungiasis (Tunga penetrans). It was later introduced in medical health and worldwide in the control of human onchocerciasis since 1988 (and has therapeutic effects on ascariasis and scabiosis).

Ivermectin acts by binding with high affinity to glutamate-gated chloride channels (GluCls) of invertebrate nerve and muscle cells causing an increase in the permeability of the cell membrane to chloride ions with hyperpolarization of the nerve or muscle cell (Cheeseman et al., 2001). This results in disruption of the neurotransmission processes regulated by the

GluCl activity in the cell. The selective activity of ivermectin and avermectin/milbymicin is

University of Ghana http://ugspace.ug.edu.gh

due to the fact that some mammals do not have GluCls and this group of drugs has low affinity for mammalian ligand-gated chloride channels. In addition, ivermectin does not readily cross the blood-brain barrier in humans but rather targets its antiparasitic activity

(Wolstenholme and Rogers, 2005). In nematodes ivermectin is found to accumulate in nerve and muscle cells when the drug is taken by the host. This makes them sensitive to the effect of the drug. Wolstenholme and Rogers (2005) showed that nematodes have multiple forms of

GluCl subunits which differ in their sensitivity to ivermectin. These have been found to be the source of differences in the sensitivity to the drug by different species of parasitic nematodes.

The ripple effect of this physiological mechanism on nematodes (disruption of the function of

GluCl) is rapid paralysis of movement and pharyngeal pumping. The worm therefore is unable to either move or feed. This effect is very drastic on microfilariae and kills them since movement and pharyngeal pumping are essential in their feeding mechanism within the host.

The adult worms may survive drug treatment because movement and pharyngeal pumping are not essential in their feeding mechanism. Rather, the drug prevents the release of microfilariae from the adult female worm. This occurs because the drug has been found to accumulate in the reproductive tissue of female worms (Li et al., 2014). Ivermectin also affects the GluCl in the muscle structure that surrounds and controls the excretory-secretory vesicles of the microfilariae (Moreno et al., 2010). Therefore, it has been proposed that ivermectin also suppresses the ability of the parasite to secrete proteins that enable evasion of the host immune system (Moreno et al., 2010). This may support the report that immunocompetence is important in the effectiveness of treating onchocerciasis with ivermectin (Ali et al., 2002). It is the abilities of ivermectin to kill the microfilariae and temporarily sterilize the adult female

University of Ghana http://ugspace.ug.edu.gh

worm that have realise the success of this drug in the control of onchocerciasis and lymphatic filariasis.

2.4.4 Responses of Onchocerca volvulus to ivermectin

Since the early 2000s there have been reports of sub-optimal responsiveness of O. volvulus to ivermectin treatment (Awadzi et al., 2004a; Awadzi et al., 2004b; Osei-Atweneboana et al.,

2007; Pion et al., 2013). These reports have been based on the persistence of microfilariae in the skin of patients who have been treated with the standard dose (150-200µg/kg body weight). The studies by Awadzi and co-workers (Awadzi et al., 2004a; Awadzi et al., 2004b) were conducted in clinically well-characterized patients and showed that the persistence of microfilariae in the skin of patients who had received several treatments was almost exclusively due to adult female O. volvulus failing to respond or responding poorly (sub- optimally) to ivermectin. Despite multiple treatments with the drug, the female worms in the

‘sub-optimal’ responders continued to produce viable/normal microfilariae compared to good responders, in whom the female worms had substantial numbers of degenerated microfilariae as assessed using embryograms (Awadzi et al., 2004a). Ivermectin was found to be effective against microfilariae from both good and poor responding female worms (Awadzi et al.,

2004a). The study by Osei-Atweneboana and co-workers, conducted in 2004–2005 was an epidemiological study aimed to assess the impact of the annual ivermectin distribution strategy in Ghana (Osei-Atweneboana et al, 2007). This study also showed that a standard dose of ivermectin is able to clear >99% of skin microfilariae one month after treatment, showing that the efficacy of ivermectin as a microfilaricide is not altered.

University of Ghana http://ugspace.ug.edu.gh

In these studies, sub-optimal responses have been measured clinically and parasitologically as: a) early recurrent pruritus in treated patients; b) a faster rate of reappearance of microfilariae in the skin following treatment than anticipated, and c) a higher proportion than expected of pre-treatment microfilaridermia levels being reached at particular timepoints after treatment. Since the microfilaricidal effect seems to remain unaffected, a possible explanation is a faster resumption of microfilarial production by the adult female worms, perhaps indicating emerging (reproductive) resistance by O. volvulus populations. However, since microfilarial levels are also associated with worm burden and high incidence (rate of acquisition of incoming worms), other explanations put forward have included programmatic factors such as low levels of therapeutic coverage and poor compliance with treatment, among others (Burnham, 2007; Cupp et al., 2007; Mackenzie, 2007; Remme et al., 2007).

A normal or good responding population, after treatment with a single standard dose of ivermectin (150-200 µg/kg), will have microfilaridermia (microfilariae in skin) almost cleared after 1-2 months (99% clearance). Females recover from the effect of the drug and begin to repopulate the skin with microfilariae. The rate of repopulation is 6% of baseline load within

6 months after treatment, 17% within 12 months and ~40% within 24 months (Basáñez et al.,

2008). These rates have been estimated for good responding worms and rates faster than these are suggested to be sub-optimal responses or indicative of a reduced effect of ivermectin on adult female worm productivity (Basáñez et al., 2008). Parasite populations with multiple treatments have been associated with faster rates of repopulation compared to naïve populations (Basáñez et al., 2008; Churcher et al., 2009). Non-responsiveness or sub-optimal responses of O. volvulus to ivermectin treatment is of great concern. This is particularly so because:

University of Ghana http://ugspace.ug.edu.gh

 Ivermectin is still the only drug used for mass treatment of onchocerciasis in many

endemic countries;

 A number of countries are aiming at eliminating the disease, stopping transmission

and/or eliminating the infection reservoir using mass treatment with ivermectin; and

 Goals have been set by the WHO to achieve elimination of onchocerciasis in selected

endemic countries by 2020 (WHO, 2012a) and in majority of endemic countries (80%)

by 2025 (WHO, 2012b).

Therefore, continuous monitoring of O. volvulus responses to ivermectin is important in order to inform control programmes on the strategies to adopt in either reducing transmission or eliminating the disease.

2.5 Mathematical Modelling of Onhcoerciasis Infection

Mathematical modelling is a mathematical process of investigating the concept and theory of infectious disease transmission to predict future occurrence in order to device possible control of any outbreak (Omade et al., 2015). There are two main types of mathematical models for epidemiological studies. These are stochastic and deterministic models. The stochastic model deals with the random study of the epidemic process using probability techniques to estimate the epidemic outcomes and measures the probability of extinction time and the size based on mean, variance and distribution. In deterministic models, the outcomes are precisely determined through known relationships among states and events, without any room for random variation. In such models, a given input will always produce the same output (Omade et al., 2015).

University of Ghana http://ugspace.ug.edu.gh

Two computer simulating models have been developed for assessing onchocerciasis control, impact of MDA and timelines for elimination. These are ONCHOSIM, developed by researchers at the Erasmus Medical Center, Erasmus University, Netherlands and

EPIONCHO, developed by researchers of Imperial College, London, UK. Some of the differences between these two models include; ONCHOSIM is a stochastic and individual- based model. Presence of infection and density (output) is at the individual level and simulation can be adjusted to run using mass treatment or selected for treatment (test and treat). The model also assumes microfilariae count as part of sampling to relate model predictions to skin-snip data (Stolk et al., 2015). On the other hand the EPIONCHO model is a deterministic and population-based model. Output is presented as mean density in population subgroups (e.g. age, sex, treatment compliance group) and prevalence obtained as a function of mean density assuming an underlying negative binomial distribution. In the latter model, chemotherapy (treatment) is based only on mass treatment and does not allow for test and treat. Sampling process and diagnostic performance of skin snipping is also not yet included in the model.

The differences in the models account for different projection timelines or required duration of mass ivermectin treatment to achieve onchocerciasis elimination (Stolk et al., 2015).

University of Ghana http://ugspace.ug.edu.gh

2.6 Onchocerciasis Vector Species in Ghana

Human onchocerciasis is transmitted by blackflies of the genus Simulium. In West Africa, the blackflies (Diptera: Simuliidae) that act as vectors of O. volvulus belong to the Simulium damnosum Theobald species complex. Within the Onchocerciasis Control Programme in

West Africa (OCP) region, where 11 West African countries were involved (Figure 2.1), nine cyto species of this complex have been identified to be transmitting the parasite with various levels of vector competence and vectorial capacity. These include: S. damnosum sensu stricto

(s.s.), S. sanctipauli, S. sirbanum, S. soubrense, S. squamosum (Enderlein), S. yahense,

S. dieguerense, S. konkourense and S. leonense (Vajime and Dunbar, 1975; Post, 1986;

Vajime, 1989; Boakye, 1993). Apart from the last three species, 6 of the sibling species are found in Ghana (namely S. damnosum s.s., S. sirbanum, S. soubrense Beffa form, S. sanctipauli Pra form, S. squamosum and S. yahense). Members of Simulium damnosum species complex differ in their geographical distribution, ecological requirement, host choice preference and vectorial capacity (Boakye et al., 1998; Adler et al., 2010; Lamberton et al.,

2014). Generally, S. damnosum s.s. and S. sirbanum are associated with the savannah bioclimatic zone of Ghana while S. sanctipauli Pra form, S. yahense, S. soubrense Beffa form, and S. squamosum are usually associated with the forest habitats, thus contributing to define the transmission dynamics and epidemiology of onchocerciasis in Ghana (Boakye et al.,

1998; Lamberton et al., 2014).

The current global and environmental changes due to rapid deforestation in southern Ghana and Togo have resulted in the immigration of savannah members of the S. damnosum to some forested regions and these have serious implications for disease epidemiology and transmission (Wilson et al., 2002; Post et al., 2013), although some of these incursions have

University of Ghana http://ugspace.ug.edu.gh

not advanced or have been described as temporary in a dynamic landscape of simuliid species distribution in southern Ghana as described by Post et al. (2013).

University of Ghana http://ugspace.ug.edu.gh

2.7 The Human Onchocerciasis Parasite (Onchocerca volvulus)

Onchocerca volvulus is a nematode. The word “Onchocerca” is broken up into two different parts. The prefix oncho- is defined as a 'tumor' in Greek and the suffix -cerca is translated into cercaria, meaning 'tail creatured' in Latin. The word volvulus means 'to twist' in Latin, which refers to how the nematode moves. The taxonomic classification of this parasite is as follows:

Domain - Eukarya

Kingdom - Animalia

Phylum - Nematoda

Class - Secernentea

Order - Spirurida

Superfamily- Filarioidea

Family - Onchocercidae/Filariidae

Genus - Onchocerca

Species - Onchocerca volvulus Leukart

Onchocerca volvulus is the only species that causes human onchocerciasis. Other Onchocerca species which cause disease in animals include O. ochengi, O. gutturosa, O. gibsoni among others (Junquera, 2015). The body of the adult worm of Onchocerca volvulus is covered with a cuticle, which is smooth and flexible but tough (Franz and Buttner, 1983). The cuticle is transversally striated forming structures like rings which are regularly spaced annulations.

The worms have a tubular digestive system with two openings; the mouth and the anus. The mouth opens at the anterior end into the esophagus which is linked to the opening at the posterior end (the anus). The worms have a nervous system but no excretory organs and no

University of Ghana http://ugspace.ug.edu.gh

circulatory system, i.e. neither a heart nor blood vessels (Junquera, 2015). They exhibit marked sexual dimorphism. The female worms are larger and longer, and measure about 33-

50 cm in length by 270-400 μm in diameter and the males measure 1.9-4.2 cm by 130-210 μm

(CDC, 2013a). O. volvulus reproduces sexually, requiring both a male and a female for reproduction. The female ovaries are large and the uterus ends in an opening called the vulva which is directly behind the anus. Male worms have their tails (spiculae) curled ventrally.

They coil around the females and use their spicules to hold the female during copulation, with the curved area on the female genital pore (Junquera, 2015). The fertilized oocytes within the females develop into morulae, horse-shoe/sausage forms, ring/coiled/pretzel forms and then to stretched or mature microfilariae as observed in other filarial nematodes (Mossinger and

Barthold, 1987). Adult females are viviparous and release hatched larvae (microfilariae) through the vulva firstly into the area of the nodule surrounding the adult worms and subsequently into the sub-cutaneous tissue of host. The microfilariae are microscopically visible and measure between 220-360 µm x 5-9 µm.

University of Ghana http://ugspace.ug.edu.gh

2.8 Candidate Proteins for Ivermectin Resistance

2.8.1 P-glycoprotein structure and function

P-glycoprotein (P-gp, permeability glycoprotein) is a trans-membrane protein and a member of the adenosine triphosphate (ATP) binding cassette (ABC) super family that functions specifically as a carrier mediated primary active efflux transporter (Amin, 2013). ABC transporters rely on energy in the form of ATP to translocate substrates (drugs) across cell membranes (Davidson et al., 2008). Several important drugs are substrates to P-gp which reduces their bioavailability or induces resistance due to the protein efflux mechanism. In humans, P-gp mediates the export of drugs from cells located in the small intestine, blood- brain barrier, hepatocytes, and kidney proximal tubule, serving as a protective function for the body against foreign substances (Wessler et al., 2013). In nematodes, P-pg has been found to induce resistance to ivermectin treatment (Kerboeuf and Guégnard, 2011).

The basic structure that defines the members of the ABC transport family consists mainly of four domains: two transmembrane domains (TMD1 and TMD2), that spans the inner and outer membrane of the cell, and two nucleotide-binding domains (NBD1 and NBD2), located in the cytoplasm (Peelman et al., 2003; Pohl et al., 2005). These domains are linked by polypeptides and arranged as TMD1-NBD1-TMD2-NBD2 or in any possible combination for proteins with two NBDs and two TMDs. The hydrophilic (cytoplasmic) NBD contains several highly conserved motifs such as Walker A, Walker B, C-motif (Klein et al., 2011; Prasad and

Goffeau, 2012) and these work together to generate the ATP binding sites.

An efflux mechanism begins first with the binding of the drug to the TMDs. This triggers a conformational change where the two NBDs in cytoplasm come closer to each other,

University of Ghana http://ugspace.ug.edu.gh

generating the ATP binding sites (Rees et al., 2009). Binding of the two ATPs sites blocks the protein in a closed state which leads to another conformational change from the inward-facing to the outward-facing conformation (Figure 2.7). The two NBDs remain bound together with their ATP while the outer leaflet part of each TMD becomes distant (open up) and release drug out of the cell (Figure 2.7). The protein then comes back to the initial inward-facing conformation by hydrolyzing ATP (Linton and Higgins, 2007; ter Beek et al., 2014).

Figure 2.7: P-glycoprotein conformational changes during drug efflux Source: http://www.cellmoloto.net/index.php/acmo/rt/printerFriendly/23955/html Accessed on 31/10/2016

University of Ghana http://ugspace.ug.edu.gh

2.8.2 Beta-tubulin structure and function

Beta-tubulins (β-tubs) are globular proteins that polymerize with alpha (α-) tubulins to form microtubules in a cell (Conde and Cáceres, 2009). These microtubules are the major component of eukaryotic and some prokaryotic cytoskeleton (Pilhofer et al., 2011). The formation of microtubules occur when the dimers of α and β tubulins bind to Guanosine-5'- triphosphate (GTP) and are assembled onto the (+) ends of microtubules while in the GTP- bound state. The β-tub subunit is exposed on the plus end of the microtubule while the α- tubulin subunit is exposed on the minus end (Conde and Cáceres, 2009). Microtubules are part of a structural network (the cytoskeleton) within the cell's cytoplasm. Their roles include mechanical/structural support, organization of the cytoplasm, intracellular transport, motility and chromosome segregation or mitosis (Stanton et al., 2011). Microtubules are capable of growing and shrinking in order to generate force, and they allow organelles and other cellular components to move through them using motor proteins (Howard and Hyman, 2003).

Microtubules are important because they serve as drug targets. Their role in mitosis makes them good targets for diverse anticancer drugs. Drugs such as ivermectin and albendazole affect the microtubules and the cytoskeletal function of cells (Aleyasin et al., 2015; Ashraf et al., 2015). Resistance to such drugs involves target site insensitivity where the structure of the target protein (e.g. beta-tubulin or microtubule) is changed and the drug is unable to bind to target site, compromising its effect (organism survives). Changes in protein structure is preceded by nucleotide change in the gene (mutations) as observed in some studies (Eng et al., 2006; Osei-Atweneboana et al., 2012). This mechanism is different from resistance using

P-pg (pumping drug out of cell) as described previously.

University of Ghana http://ugspace.ug.edu.gh

CHAPTER THREE

GENERAL METHODOLOGY

3.1 Ethical Approval

Ethical approval was obtained from the ethics review committees of the Noguchi Memorial

Institute for Medical Research, Ghana (NMIMR-IRB CPN 032110-11), Council for Scientific and Industrial Research, Ghana (RPN 003/CSIR-IRB/2011 and RPN 003/CSIR-IRB/2013)

[see Appendix I], the Ghana Health Service [GHS ERC 04_3_11] and Imperial College,

London, Research and Ethics Committee [ICREC_11_2_4].

3.2 Study Sites

Samples were collected from 10 selected endemic communities within the Brong-Ahafo and

Northern Regions (savannah areas) of Ghana, areas known to be associated with severe cases of onchocerciasis (blinding form) [Figure 3.1]. The Brong-Ahafo Region (BA) is located on

7°45'0" N and 1°30'0" E at an altitude of 170 m above sea level. It is situated in the mid- western part of Ghana between the Ashanti and Northern Regions. It shares common boundaries with the Northern Region to the north, Ashanti and Western Regions to the south, the Volta Region to the east and the Eastern Region to the south east. It has an international boundary to the west which it shares with Ivory Coast. Temperature in the region is generally high, averaging over 23.9°C (21 - 34°C) throughout the year and relative humidity in the region is also high, averaging over 75% (62 - 97%) throughout- the year. Communities selected from BA were Kyingakrom and New-Longoro (Kintampo District), Asubende,

Baaya, Ohiampe and Senyase (Pru District) [Figure 3.1]. These communities lie close to rivers which are source of breeding sites for the blackflies. Asubende, Baaya, Ohiampe and

University of Ghana http://ugspace.ug.edu.gh

Senyase lie close to the Pru River. Kyingakrom and New-Longoro lie close to the Lower

Black Volta (LBV) River. Majority of the people living in these communities are farmers involved in crops and rearing of livestock especially goats, sheep and cattle. Occasionally some engage in fishing since most of them live close to the rivers.

The Northern Region (NR) is located on 9°30'N and 1°00'W and shares borders on the north with the Upper West Region and the Upper East Region, on the east by the eastern Ghana-

Togo border, on the south by the Black Volta River and the Volta Region and on the west by the western Ghana-Ivory Coast border. It is the largest region in the country and much drier than southern areas of Ghana, due to its proximity to the Sahel, and the Sahara. The vegetation consists predominantly of grassland, especially savannah with clusters of drought- resistant trees. The average annual rainfall ranges from 750 to 1050 mm. The temperatures can vary between 14°C (59°F) at night and 40°C (104°F) during the day. Communities selected from the NR were Jagbengbendo, Takumdo, Wiae (Kpandai District) and

Agborlekame 1 (Bole District) [Figure 3.1]. Agborlekame 1 lies close to the LBV River whiles Jagbengbendo, Takumdo and Wiae district lie close to the Daka River. Majority are farmers involved in yam cultivation and rearing of livestock especially goats, sheep and cattle.

They are also involved in occasional fishing.

University of Ghana http://ugspace.ug.edu.gh

Figure 3.1: Map of study sites in Brong-Ahafo and Northern Regions of Ghana. Prepared by MAPTECH SYSTEMS, P.M.B. 1AF, Adenta Flats, Accra.

University of Ghana http://ugspace.ug.edu.gh

3.3 Study Design and Parasitological Methods

The 10 selected study communities are part of the onchocerciasis endemic communities selected by the Ghana Health Service (GHS) for mass biannual treatments with ivermectin, which started in July 2010. These communities had history of previous treatments with good treatment records (Osei-Atweneboana et al., 2007). Some of these communities had also been previously implicated as sub-optimally responding to ivermectin treatment (Awadzi et al.,

2004a; Awadzi et al., 2004b; Osei-Atweneboana et al., 2007; Churcher et al., 2009) and had received between 15 and 24 rounds of annual ivermectin treatment prior to this study. Adults who were ≥20 years were selected from different households during voluntary participation within each community. The criteria for selection was based on the age group used by the

Onchocerciasis Control Programme (OCP) to assess the impact of vector control in endemic communities (Remme et al., 1986). The objectives and schedules of the study were explained to every individual, and those who agreed to participate signed a consent form (see Appendix

I).

The study design and timelines are illustrated in Figure 3.2. Before the first round of biannual treatment, in July 2010, a community-wide skin snipping was performed on 956 consenting adult participants to assess the community prevalence and intensity of microfilariae. Skin snips were taken from the areas of the body around the two iliac crests using a 2 mm Holth- type corneoscleral punch (Stephens Instruments, USA). The skin snips were placed in a 0.9% sterile saline (or physiological saline) solution and, using an inverted microscope, the microfilariae were counted after 30 minutes of incubation. Skin snips both positive and negative with microfilariae were further incubated until 24 hours and re-counted. All 956 participants after skin snipping were treated with ivermectin (150 µg/kg body weight) by the

University of Ghana http://ugspace.ug.edu.gh

Community Ivermectin Distributors (CIDs) [day 0/baseline]. After approximately six months, in January 2011, another round of skin snipping was taken from all the 956 participants (both positives and negatives at baseline) immediately before administering the second round of ivermectin treatment. A total of 217 participants who were microfilaria positive at baseline were followed-up for further skin snips in April 2011 and July 2011, i.e. 3 and 6 months after the second ivermectin treatment round (Table 3.1). The July 2011 skin snipping was immediately followed by the third round of ivermectin, administered to the entire population.

In November 2011, a total of 106 follow-up participants with palpable nodules who consented through written documentation had their nodules removed under local anesthesia in the closest health facility to the community (by a surgeon). This was to assess the drug effect on the adult female worms approximately three months after treatment by embryogram analysis.

The communities received the 4th, 5th and in some cases 6th rounds of ivermectin treatment approximately every six months by the CID as part of the regular Neglected Tropical Disease

Control Programme (NTDCP) activities of the GHS. It must however be noted that some communities did not implement every scheduled round of treatment. Just before the last treatment round ie. 5th or 6th round for some communities, a final community-wide round of skin snips was taken in June 2013 for most communities (Figure 3.2). Compliance of all treatments in this study was ensured by direct observation (WHO, 1998). Reported treatment coverage data was obtained from the NTDCP and analysed as part of this study.

University of Ghana http://ugspace.ug.edu.gh

3.4 Isolation of Worms

Nodules taken from participants in November 2011 were transported to the laboratory in liquid nitrogen. These were digested using the collagenase techniques to isolate the male and female worms. Worms were cleaned with sterile 0.9% saline and the intact female worms used for embryogram analysis to determine the drug effect on stretched microfilariae within the uterus (see Chapter 5 for more details). Worms were classified into good response (all microfilariae in uterus degenerated or no stretched microfilariae) and poor response (>80% or all normal stretched microfilariae present in uterus) parasites.

3.5 Molecular Studies

Deoxyribonucleic acid (DNA) was extracted from good and poor responding worms for amplification using primers specific for one region within the beta-tubulin (β-tub) gene and six regions within the P-glycoprotein (P-gp) gene using the Polymerase Chain Reaction

(PCR). The amplified PCR products were sequenced and multiple sequence alignment done comparing results with the reference sequences of β-tub and P-gp. Any single nucleotide polymorphism (SNP) associated with good or poor/sub-optimal response to ivermectin treatment was determined (see Chapter 6 for more details).

University of Ghana http://ugspace.ug.edu.gh

Figure 3.2: Schematic timelines of participants (Study design) Figure shows the onset of a biannual treatment strategy in 10 communities in Ghana. Participants 1–6 represent the individuals from whom skin snips were taken in July 2010, just before the first round of biannual ivermectin treatment, and 6 months later in January 2011, just before the second round of biannual ivermectin treatment. Participants 1–5 were positive for microfilariae in July 2010 and hence were included in the cohort of 217 individuals for evaluating rates of skin microfilarial repopulation. Participants 1–4 represent individuals who were microfilaria positive in January 2011, with participants 1 and 2 successfully followed up and skin snipped in April 2011 and again in July 2011, just before the third round of biannual ivermectin treatment. Participants 1, 3, 4, and 6 represent participants who agreed to be skin snipped for a final time in June 2013, just before the final round of treatments. The months given on the timeline are the modal months of treatment activity among the 10 communities, but there is significant variation in the months and exact dates, especially for the biannual treatments given after July 2011 (see Chapter 4, Figures 4.1 and 4.2 for exact dates plotted)

University of Ghana http://ugspace.ug.edu.gh

Table 3.1: Longitudinal cohorts of participants in the 10 Ghanaian communities who were followed up and skin snipped over the first two rounds of biannual treatment with ivermectin

Community Month and year (months since preceding round of treatment) July 2010a January 2011b April 2011 July 2011

Agborlekame 1 63 27 23 20 Asubende 34 9 8 9 Baaya 129 1 1 1 Jagbengbendo 107 50 47 46 Kyingakrom 82 14 12 12 New-Longoro 126 17 13 15 Ohiampe 85 5 5 4 Senyase 64 8 6 7 Takumdo 108 50 48 44 Wiae 158 26 23 24 Total 956 217 186 182 aOnly participants positive for microfilariae were followed up in January 2011; bonly participants positive for microfilariae followed up in April 2011 and July 2011.

University of Ghana http://ugspace.ug.edu.gh

CHAPTER FOUR

RESPONSES IN ONCHOCERCIASIS PATIENTS AFTER BIANNUAL IVERMECTIN TREATMENT IN GHANA

4.1 Introduction

Ghana was one of the first countries to introduce ivermectin mass treatment to control onchocerciasis in 1987 after the drug was licensed for use in humans (Crump and Omura,

2011). The first community trials were conducted in Asubende, an endemic community within the Brong-Ahafo Region of Ghana by the Onchocerciasis Control Programme in West

Africa (OCP, 1974–2002) [Alley et al., 1994].

Ivermectin kills Onchocerca volvulus microfilariae, which are the larval progeny of adult worms (transmissible form of parasite). Ivermectin also temporarily sterilizes female worms such that numbers of microfilariae remain heavily suppressed for at least three months following treatment (Basáñez et al., 2008). Subsequently, females regain fertility and microfilariae repopulate the skin. Hence, ivermectin can only control onchocercal disease which is predominantly caused by chronic infestation by microfilariae within the skin and ocular tissue (Brattig, 2004), if the drug is given at regular intervals. In an endemic community, if microfilariae are suppressed to such an extent that transmission is interrupted, and this is maintained for at least as long as the 10 year average lifespan of the adult worm

(Plaisier et al., 1991), it is possible to eliminate the infection. Mass treatments with ivermectin have successfully eliminated onchocerciasis from three foci in Mali and Senegal (Traore et al., 2012) and two in Nigeria (Tekle et al., 2012), as well as in Mexico (Rodriguez-Perez et al., 2015), Colombia (West et al., 2013), Ecuador (Lovato et al., 2014), northern Venezuela

University of Ghana http://ugspace.ug.edu.gh

(Convit et al., 2013) and Guatemala (Sauerbrey et al., 2018). The strategy used in Latin

America was mostly biannual treatment.

Despite years of ivermectin treatment, onchocerciasis still affects individuals or communities within 66 districts in Ghana (GHS, 2007), approximately 3.2 million people remain at risk of infection (Taylor et al., 2009). The resilience of onchocerciasis to control is likely, in part, caused by the documented poor responses to ivermectin of people in several Ghanaian communities; adult female worms in these individuals tend to produce more viable microfilariae compared to females from normal responders (Awadzi et al., 2004a; Awadzi et al., 2004b). In a community of normally responding individuals, one expects numbers of microfilariae to be at about 10% of their pretreatment numbers six months after treatment, and at about 20% of their pretreatment numbers one year after treatment (Basáñez et al., 2008;

Katabarwa et al., 2014). In contrast, sub-optimally or atypically responding communities repopulation rates six months after treatment have been observed at over 50% of pretreatment values (Osei-Atweneboana et al., 2007). Moreover, some of these communities are those that have been treated with the most rounds of ivermectin (Churcher et al., 2009).

In 2010, in response to the persistence of onchocerciasis in Ghana, the Neglected Tropical

Diseases Control Programme (NTDCP) of the Ghana Health Services (GHS) adopted a biannual (twice per year) treatment strategy in 44 out of the 74 re-demarcated endemic districts (Turner et al., 2013a).

The purpose of this chapter is to estimate the current prevalence and microfilarial loads in 10 sentinel communities of the NTDCP (some previously identified as responding sub-optimally

University of Ghana http://ugspace.ug.edu.gh

to ivermectin) before and after four or five rounds of biannual treatment. The study also assessed the rates of microfilarial repopulation in cohorts of individuals from each community followed-up at six months and one year after treatment with ivermectin, comparing skin repopulation rates with community endemicity and the number of years of prior ivermectin treatment.

4.2 Methods

This was a follow-up study from July 2010 to June 2013 (see schematic timeline and study design, Figure 3.1 in Chapter 3). A total of 956 participants sampled at baseline in July 2010 from different households were skin snipped and treated. The 956 participants were again snipped and treated at six months (January 2011). A cohort of 217 participants who were positive at baseline were followed up on January 2011, April 2011 and July 2011 to assess microfilariae loads and determine repopulation rates. Participants in the 10 communities were sampled finally during a community survey in June 2013.

4.2.1 Data analyses

The community microfilarial load (CMFL) as described (Remme et al., 1986) and the community microfilarial prevalence (CMFP) were used as indicators of the average intensity and prevalence of O. volvulus infection in the adult (≥ 20 years) population respectively. The

CMFL is defined as the geometric mean number of microfilariae (including zero counts) per skin snip (in people aged ≥20 years) and the CMFP is defined as the prevalence of microfilariae (in people aged ≥20 years) in the community. These were estimated before treatment in July 2010 (baseline), January 2011 and March 2013 or June 2013. Details of the

CMFL and the CMFP calculations have been given in Appendix II A. Coverage was

University of Ghana http://ugspace.ug.edu.gh

calculated using treatment and census data provided by the Community Ivermectin

Distributors (CIDs) to the NTDCP and it refers to the percentage treated out of the total population at risk (including children less than 5 years or with height less than 90 cm, pregnant and lactating mothers as well as sick individuals). The community treatment coverage at all treatment rounds were obtained from the NTDCP, GHS to facilitate interpretation of the CMFL and CMFP values obtained throughout the study period.

The mean number of microfilariae per skin snip (mf/ss) in the longitudinal cohort of 217 individuals (Table 3.1, Figure 4.2) was estimated using log-linear marginal regression models

(Diggle et al., 2013) and adjusted for community, participant age, and sex. Two models were used for these analyses (Table 4.1). In both models, repopulation rates varied among communities, but in the first model (Model 1A and 1B, Table 4.1), repopulation rates varied between the 2 consecutive repopulation periods (January 2011 and July 2011), whereas in the second model (Model 2, Table 4.1), a single community-specific repopulation rate was estimated, combining both repopulation periods. Microfilarial repopulation rate is defined as the mean number of microfilariae expressed as a percentage of the previous mean (before treatment with ivermectin). This captures how quickly microfilariae reappear in the skin between consecutive treatment rounds. The mathematical details have been given in

Appendix II C.

The repopulation rates observed from participants were compared with the model-fitted

(Model 1A, Table 4.1) mean number of mf/ss within males aged 21-40 which was used as the reference demographic stratum. This was used as a reference stratum because males within this age group formed the majority of the participants. Comparison was done to assess the

University of Ghana http://ugspace.ug.edu.gh

performance of the model. Since there were no data collected in October 2010 (3 months after the first round of biannual treatment), the mean numbers of mf/ss at that time point were predicted using the marginal regression model (Model 2 in Table 4.1). This model treated the time since the preceding ivermectin treatment as a continuous variable. The predicted values in October 2010, were provided to envisage the likely dynamics in mean numbers of mf/ss during the first 6-month repopulation period.

The microfilarial repopulation rates of all communities were compared with Takumdo which was used as the reference community (least treatment years – 15 years), since all endemic communities were under treatment. Therefore no naïve community was available. The repopulation rates were explored graphically to see how they correlated with the number of years of annual ivermectin treatment and CMFL just before the start of biannual treatment. A

95% confidence interval associated with all estimates were calculated using a numerical bootstrap resampling method (Davison and Hinkley, 1996) and significance observed at

P < 0.05. Details of the numerical bootstrap resampling method have been given in

Appendix II B.

University of Ghana http://ugspace.ug.edu.gh

Table 4.1: Key features of the Log-Linear Marginal Regression Models used to describe the observed microfilarial counts in the longitudinal cohort Type Variant Key features

Model 1 A and B  Response/outcome variable defined by individual microfilarial counts  Modelled mean number of microfilariae per participant adjusted for the covariates age group (18-20, 21-40, 41-60 and 61-80), sex and community  Microfilarial repopulation rates permitted to vary among communities and between repopulation periods by including sampling time as a categorical covariate interacting with community B  Microfilarial repopulation rates adjusted by exact number of days since preceding round of ivermectin treatment yielding standardized repopulation rates (e.g. 6-month repopulation rates) in each community Model 2  Response/outcome variable defined by individual microfilarial counts  Modelled mean number of microfilariae per participant adjusted for the covariates age group (18-20, 21-40, 41-60 and 61-80), sex and community  A single microfilarial repopulation rate estimated for each community, combining information from both repopulation periods, by including sampling time as a continuous covariate (defined as days since preceding ivermectin treatment) interacting with community  Additive, community-wide adjustments for potentially different repopulation rates between 2 repopulation periods Model variant not applicable for Model 2. See Appendix II C for details and explanation of model

University of Ghana http://ugspace.ug.edu.gh

4.3 Results

The ivermectin treatment coverage within the 10 communities ranged from 42 – 93% between

July 2010 to June 2013 (Figures 4.1 and 4.2). Out of the 956 participants enrolled at pretreatment (baseline), 22.7% were positive for skin microfilariae (0.5 – 38%) and 26.8% positive for palpable nodules (12 – 43%) [Table 4.2]. The CMFL at pretreatment ranged from

0 – 1.57 mf/snip with minimum observed in Baaya and maximum in Takumdo (Table 4.2).

Agborlekame 1, Jagbengbendo and Takumdo showed high nodule and microfilariae prevalence among the 10 communities. At the end of study in 2013, 12.5% (N = 582) were positive for skin microfilariae (2.4 – 25.7%). The CMFL ranged from 0.01 – 0.58 mf/snip.

Summary of the prevalence and intensity of onchocerciasis in the communities studied are given in Table 4.2.

Table 4.2: Prevalence and intensity of onchocerciasis infection at baseline and end of study Years of Number % Nodule % Mf CMFL at % Mf CMFL at Community annual examined prevalence prevalence pre- prevalence end of ivermectin at pre- at pretreatment at end of study - treatment treatment treatment (mf/snip) study 2013 before (2013) (mf/snip) study Agborlekame 1 24 63 42.9 34.6 1.04 17.6 0.31 Asubende 24 34 29.4 20.1 0.33 15.5 0.17 Bayaa 23 129 12.4 0.5 0.00 2.4 0.03 Jagbengbendo 20 107 41.1 38.3 1.09 25.7 0.58 Kyingakrom 23 82 25.6 12.6 0.32 12.5 0.35 New-Longoro 23 126 15.9 9.8 0.20 2.5 0.01 Ohiampe 23 85 25.9 4.2 0.09 4.9 0.15 Senyase 23 64 25.0 9.1 0.11 3.4 0.04 Takumdo 15 108 37.0 37.8 1.57 6.6 0.17 Wiae 18 158 25.3 12.1 0.35 5.9 0.10 Data on years of annual ivermectin treatment was obtained from the Neglected Tropical Diseases Control Programme of the Ghana Health Service

University of Ghana http://ugspace.ug.edu.gh

4.3.1 Trends in community infection

There was a general decline in both CMFP and CMFL between treatment round 1, in July

2010 and the end of study in June 2013, though the decline in CMFL was greater than that observed for CMFP. Figures 4.1 and 4.2 also show dates of treatment and coverage. In

Agborlekame 1, Jagbengbendo and Takumdo, the decline in the CMFP was more apparent than the rest of the communities (Figure 4.1). Yet, the mean reduction in these three communities was not statistically significant from the mean reduction of the rest (P = 0.57).

The impact of the first round of biannual treatment appeared greater than that of subsequent rounds, as demonstrated by the generally more marked decline in CMFL and CMFP (Figures

4.1 and 4.2) between treatment round 1, in July 2010, and treatment round 2, in January 2011, compared to that between treatment round 2 and the final assessment of infections levels in

March 2013 or June 2013 (Figures 4.1 and 4.2). This trend is particularly apparent in

Asubende, Jagbengbendo, New-Longoro, Senyase and Wiae, but less pronounced in

Agborlekame 1 and Takumdo. In Ohiampe, community infection levels were greater in June

2013 than in July 2010, despite 4 rounds of treatment (one treatment round was missed in the first quarter of 2013). Treatment impact on CMFL was generally greater than CMFP, though mean reduction in CMFL was not statistically significant from mean reduction of CMFP

(P = 0.09).

University of Ghana http://ugspace.ug.edu.gh

Figure 4.1: Trends in community microfilarial prevalence (CMFP) in the 10 communities after a biannual ivermectin treatment. In each panel, data points (black squares) show the CMFP from the onset of a biannual ivermectin treatment strategy and the trends are represented by the dashed lines. The vertical lines are error bars which indicate 95% confidence intervals calculated using a non-parametric bootstrap technique (see Appendix II B). The arrows of the coloured triangles pointing downwards indicate specific dates when mass treatment with ivermectin was distributed and the percentages above the triangles indicate the therapeutic coverage in the whole community.

University of Ghana http://ugspace.ug.edu.gh

Figure 4.2: Trends in community microfilarial loads (CMFLs) in the 10 communities after a biannual ivermectin treatment. In each panel, data points (black squares) show the CMFLs from the onset of a biannual ivermectin treatment strategy and the trends are represented by the dashed lines. The vertical lines are error bars which indicate 95% confidence intervals calculated using a non-parametric bootstrap technique (see Appendix II B). The arrows of the coloured triangles pointing downwards indicate specific dates when mass treatment with ivermectin was distributed and the percentages above the triangles indicate the therapeutic coverage in the whole community.

University of Ghana http://ugspace.ug.edu.gh

4.3.2 Trends in microfilarial repopulation

In general, mean numbers of microfilariae per stratum were lower after the second repopulation period than after the first repopulation period (Figure 4.3). The mean numbers of microfilariae per stratum estimated in January 2011, six months after the start of biannual treatment, appeared generally high i.e. about 50% of the estimates for July 2010 (Figure 4.3); one expects microfilariae to reach only about 10% of their pretreatment population level after six months (Basáñez et al., 2008). The 50% six months repopulation rate was statistically different from the expected 10% six months repopulation rate (P < 0.0001).

4.3.3 Microfilarial repopulation rates

The rates of microfilarial repopulation were generally high, typically approximately 50% during the first period of repopulation, and similar, though somewhat more variable, during the second repopulation period (Figure 4.4). The repopulation rates in Asubende and

Kyingakrom after the second round of ivermectin treatment statistically were found to be significantly higher than that in the reference community, Takumdo (Figure 4.4). Estimating the repopulation rates in July 2011 as a percentage of microfilarial loads in July 2010

(baseline), the repopulation rates in Asubende, Kyingarom and New-Longoro were found to be statistically significantly (P < 0.05) higher than Takumdo (Figure 4.5A). There was no association between the relative rate of microfilarial repopulation and the number of years of annual treatments with ivermectin before the start of study (Figure 4.5B) [model coefficients not significantly different from zero; P = 0.84]. No association was also observed between the relative rate of microfilarial repopulation and the CMFL at baseline before the first biannual treatment (Figure 4.5C) [model coefficients not significantly different from zero; P = 0.63].

University of Ghana http://ugspace.ug.edu.gh

Figure 4.3: Trends in mean numbers of microfilariae per participant in 9 communities after a biannual ivermectin treatment. In each panel, data points (squares) represent observed mean microfilarial loads (per community) in males within the age group 21–40 years and the solid vertical lines correspond to 95% confidence interval showing the range of possible individual values; estimated using bootstrapping (see Appendix II B). The solid continuous line join fitted values (males of age group 21-40 years) which were estimated using the marginal regression model (Model 1A in Table 4.1, also see Appendix II C). In October 2010, values represent predicted estimates also generated from the marginal regression model that treated the time since the preceding ivermectin treatment as a continuous variable (Model 2 in Table 4.1, also see Appendix II C) and assumes that (hypothetical) microfilarial sampling took place midway between the July 2010 and January 2011 sampling times. The two dashed lines join points corresponding to 95% confidence interval of the fitted estimates (calculated using robust sandwich estimators of coefficient standard errors, Appendix II D). Baaya was not included because only 1 participant was microfilaria positive and followed up in this community (Table 3.1, Chapter 3), leading to very large associated estimates of uncertainty. The arrows of the triangles pointing downwards indicate when ivermectin was administered to the study participants

University of Ghana http://ugspace.ug.edu.gh

Figure 4.4: Six-month microfilarial repopulation rates in 9 communities after a biannual ivermectin treatment. The filled and open data points (squares) represent, respectively, estimated mean microfilarial loads 6 months after the first and second round of ivermectin treatment, expressed as a percentage of the microfilarial loads just before the previous round of ivermectin treatment. These estimates were derived from Model 1B in Table 4.1. Vertical lines indicate 95% confidence interval, calculated using robust sandwich estimators of coefficient standard errors (Appendix II D). The P values compared the rate of repopulation with the reference community, Takumdo: ***P < 0.001; **P < 0.01. The 6-month microfilarial repopulation rate for Baaya was not included because only 1 participant was microfilaria positive and followed up in this community (Table 3.1, Chapter 3), leading to very large associated estimates of uncertainty.

University of Ghana http://ugspace.ug.edu.gh

Figure 4.5: Relative 6-month microfilarial repopulation rates in 9 communities over the first 2 rounds of a biannual ivermectin treatment strategy. In each panel, data points represent the estimated relative (multiplicative) 6-month microfilarial repopulation in each community compared with Takumdo. Six-month repopulation rates are defined as mean microfilarial loads 6 months after a round of ivermectin treatment, expressed as a percentage of the microfilarial loads just before the previous treatment round. Estimates are derived from Model 2 (Table 4.1): A: Estimates were plotted side-by-side for the different communities. The 6-month microfilarial repopulation rate in Baaya was not included because only 1 participant was microfilaria positive (Table 3.1, Chapter 3), leading to very large associated estimates of uncertainty. B: Estimates were plotted against the number of years of annual ivermectin treatment before the biannual strategy. C: Estimates were plotted against community microfilarial load (CMFL) at baseline before the first biannual ivermectin treatment. The vertical lines are 95% confidence interval, calculated using robust sandwich estimators of coefficient standard errors (Appendix II D). The solid horizontal lines in (C) indicate 95% confidence intervals associated with the estimated CMFL, calculated using a numerical bootstrap resampling method (Appendix II B). *P < 0.05, comparing with the reference community, Takumdo.

University of Ghana http://ugspace.ug.edu.gh

4.4 Discussion

Transmission of onchocerciasis in Ghana continues despite the long-standing and large-scale

(antivectorial and antiparasitic) interventions implemented over the past four decades

(Lamberton et al., 2015). The disease was earmarked for elimination as a public health problem by 2015 (Taylor et al., 2009), yet, transmission exists within some hotspots (Osei-

Atweneboana et al., 2007; Lamberton et al., 2014) and there are reports of O. volvulus microfilariae repopulating in the skin of patients faster than expected following treatment with ivermectin (Osei-Atweneboana et al., 2007; Churcher et al., 2009; Pion et al., 2013). In response to this challenge, the Ghana Health Services (GHS) implemented a biannual mass ivermectin treatment in many endemic communities in 2010 (Turner et al., 2013a). The aim of this study was to show the trends in community-wide infection with O. volvulus in 10 communities over the first 3 years of this biannual strategy and to evaluate the rates of microfilarial repopulation in cohorts of participants over the first 2 rounds of treatment in

Ghana.

Before the onset of this biannual strategy in 2010, a similar study had been carried out between 2004 and 2005 in these same communities. At the time of that study, these communities had received between 10–18 annual mass ivermectin treatments in Ghana (Osei-

Atweneboana et al., 2007). Comparing infection levels in the 2004 – 2005 study with this recent study (July 2010) showed that the intervening 6 years of annual ivermectin treatment reduced CMFLs generally by at least 50% and CMFP by at least 12% in majority of the communities (Appendices III and IV). Infections were further reduced by June 2013, after 3 years of the biannual treatment. By this time, reductions in CMFL were >40% in most communities and the CMFP was significantly <10% in 5 of 10 communities (Appendices III

University of Ghana http://ugspace.ug.edu.gh

and IV). The biannual treatment strategy indeed had a positive impact, though in communities where microfilarial prevalence was >10%, or >20% as in Jagbenbendo, it is difficult to ascertain whether these communities are on the track to reach elimination. Whether biannual treatments will ultimately be sufficient to eliminate infection by the WHO set target in

2020 / 2025 (WHO, 2012a) will depend on local transmission and programmatic conditions, especially on the intensity of blackfly biting (Lamberton et al., 2014; Lamberton et al., 2015) and the sustainability of high levels of treatment coverage and compliance (Turner et al.,

2013b; Turner et al., 2014b).

The mean numbers of microfilariae per stratum were lower after the second repopulation period than after the first repopulation period in most of the communities because microfilariae are unable to repopulate the skin completely in six months before their numbers are further suppressed by another round of treatment (Figure 4.3). The six-month rates of repopulation estimated (Figure 4.4) were generally around 50% which are high compared with the expected 10% from parasite populations predominantly naive to ivermectin (Basáñez et al., 2008). They are also higher than those estimated from some of the same communities in the previous study (Osei-Atweneboana et al., 2007), which were typically <30% (Appendix

V). Some of these discrepancies may probably be because the 10% value (estimated previously from these communities) was based on geometric means, which are not strictly comparable with the model-derived repopulation rates presented in this study (arithmetic means). In addition, the sampling scheme employed in this study (and also in the previous study) followed up only participants who were positive for microfilariae at recruitment. This protocol ensures that only people infected with O. volvulus are repeatedly skin snipped, increasing the efficiency of sampling even when the prevalence of infection is low. Such a

University of Ghana http://ugspace.ug.edu.gh

procedure can potentially introduce sampling biases because the sensitivity of skin snipping declines with decreasing infection intensity (Taylor et al., 1989b). Therefore, participants with less intense infections at baseline were more likely to be erroneously deemed uninfected and not followed up. This probably biased repopulation rates upward because more intensely infected people would have more microfilariae after a period of repopulation than those with less intense infections. Notwithstanding, the 3 communities with the highest repopulation rates over the 2 repopulation periods (Asubende, Kyingakrom and New-Longoro) have been previously implicated as responding sub-optimally to ivermectin (Osei-Atweneboana et al.,

2007; Churcher et al., 2009; Osei-Atweneboana et al., 2011). The cause of these observations cannot be determined from this study. However, previous suggestions that faster rates of skin repopulation by microfilariae might result from a sudden increase in new infections between treatment rounds—perhaps due to programmatic deficiencies in coverage and compliance

(Cupp et al., 2007; Remme et al., 2007)—are difficult to reconcile with the generally high levels of therapeutic coverage observed throughout this study (Figures 4.1 and 4.2). It is more likely that transmission has been declining since the onset of biannual ivermectin treatment in

July 2010, as evidenced by the generally falling CMFL (Figure 4.2), but the resilience of community infection levels to biannual distribution in a community like Kyingakrom is worth noting.

The resilience to treatment may be genetically driven. Previous analyses comparing allele frequencies among adult female O. volvulus in individuals treated multiple times and ivermectin naive populations in Ghana and Cameroon identified selection of P-glycoprotein and beta-tubulin (β-tub) genes, both associated with resistance to ivermectin in helminth infections of livestock (Eng et al., 2006; Bourguinat et al., 2007). Moreover, a genetic

University of Ghana http://ugspace.ug.edu.gh

analysis of the entire region of the β-tub gene extracted from worms infecting people from

Kyingakrom (a consistently implicated sub-optimally responding community) has identified statistically significantly higher frequencies of 6 single-nucleotide polymorphisms (Osei-

Atweneboana et al., 2012). It is not well understood how the phenotypic response of individual worms relates to these genetic changes. Worms collected from sub-optimally responding communities have been associated with higher fertility than worms from putatively normally responding communities (Osei-Atweneboana et al., 2012), possibly indicative of a faster resumption of fertility following exposure to ivermectin (Pion et al.,

2013). Another study suggests that selection driven by exposure to ivermectin is rather associated with a pleiotropic fitness cost of decreased fertility, so perhaps resistant worms may resume production of microfilariae more rapidly than their susceptible counterparts, but ultimately have less reproductive potential.

The microfilarial repopulation rates in this study were based on average, community estimates, adjusted for individual (host) characteristics such as age and sex. This is consistent with the inferential basis of previous, more descriptive analyses of data from some of the same communities studied here (Osei-Atweneboana et al., 2007; Osei-Atweneboana et al.,

2011). Particularly for these well-studied and relatively small communities, many of the same individuals have probably repeatedly participated in the epidemiological studies undertaken over the last 15 years. Therefore, future analyses should focus on estimating drug responses at the individual level (Walker et al., 2014; Coffeng et al., 2014b). It is possible that certain individuals, rather than entire communities, are consistently responding poorly to ivermectin

(and influencing the community-wide response). Poor individual responses to treatment might be caused by host-related factors or drug-tolerant parasites (given the long lifespan of adult O.

University of Ghana http://ugspace.ug.edu.gh

volvulus). The biannual ivermectin treatment strategy is markedly reducing O. volvulus infection levels in Ghana. However, despite high and sustained therapeutic coverage, sub- optimal responses to ivermectin persist in previously implicated communities. Whether this is caused by drug-tolerant or resistant parasites, or by host-related factors, remains unclear.

There is the need to scale up interventions such as the use of alternative drugs (e.g. doxycycline or moxidectin) [Turner et al., 2015; Walker et al., 2015], vector control, active surveillance etc. to meet the WHO 2020 / 2025 set target.

University of Ghana http://ugspace.ug.edu.gh

CHAPTER FIVE

EMBRYOSTATIC EFFECT OF IVERMECTIN ON ONCHOCERCA VOLVULUS ADULT FEMALE WORM

5.1 Introduction

Transmission of onchocerciasis begins with an infective bite of the blackfly (Simulium damnosum) vector carrying L3 stage larvae of Onchocerca volvulus. After infection, the L3 migrate to the epidermis of the host where they moult via L4 and L5 stages to become adult worms and eventually after about nine to twelve months become reproductively mature

(Duke, 1980; Prost, 1980). The adult worms are found in nodules (onchocercomata) which harbour male and female worms intertwined with each other. In the presence of at least one male worm, females in the nodule are fertilized (there are 4-5 reproductive cycles per year) and continue to produce their progenies (microfilariae). The fertilized oocytes within the females develop into morulae, horse-shoe/sausage forms, ring/coiled/pretzel forms and then to the stretched or matured microfilariae as observed in all filarial worms (Mossinger and

Barthold, 1987). This process may take between three to four weeks (Schulz-Key and Karam,

1986). After their full development, the microfilariae are released into the host where they migrate through the nodules in the sub-cutaneous tissue to the epidermis and lymphatics of the skin, where they are mostly located till they are picked by the blackfly through a blood meal. The mature female worm, when fertilized, can produce about 1,000 stretched microfilariae per day (Engelbrecht and Schulz Key, 1984).

Currently onchocerciasis is treated with ivermectin, which kills the microfilariae in the skin at a standard dose of 150-200 µg/kg body weight. However, it does not kill the adult worm,

University of Ghana http://ugspace.ug.edu.gh

unless given repeatedly (biannual or quarterly frequencies) or at an increased dose of 400-

800 µg/kg body weight (Gardon et al., 2002; Cupp et al., 2004)). Nonetheless, the drug has a marked embryostatic effect on the adult female worm by preventing the release of stretched/matured microfilariae from its uterus. These normal stretched microfilariae (Figure

5.4A) as a result accumulate in the uterus, become vacuolized and then degenerate within a few weeks (Schulz-Key and Karam, 1986; Awadzi et al., 2004a; Awadzi et al., 2004b).

Vacuoles appear as tiny cavities or spaces within the microfilariae (Figure 5.4B) and then their internal contents detach from the inner walls showing early signs of degenerating microfilariae. The microfilariae become darkened and beaded at the late stages of degeneration (Figure 5.4C) and then reabsorbed into the uterus (Schulz-Key and Karam,

1986; Tielsch and Beeche, 2004). This effect persist in subsequent reproductive cycles of the female worm up to 9 months (Schulz-Key and Karam, 1986). Although multiple doses of ivermectin can be macrofilaricidal (Gardon et al., 2002), the death of adult worms may also occur as a result of aging, which on the average is ten years (the lifespan of the worm). These are manifested through the calcification of nodules and/or worms (Albiez et al., 1984; Albiez,

1985).

Apart from ivermectin treatment, other factors such as the age of adult female worms may affect the production of microfilariae (Schulz-Key and Karam, 1986). The frequency of reproductive cycles decreases with advancing age of the worms, resulting in the reduction of net microfilarial production and more prolonged intervals between reproduction cycles.

Unlike other filarial worms, O. volvulus requires insemination by adult males for each reproductive cycle. Therefore as the worms advance in age, the older parasite populations

University of Ghana http://ugspace.ug.edu.gh

may have female worms fertilized by males but with empty uteri (Schulz-Key and Karam,

1986; Chavasse et al., 1993).

Ivermectin is still efficient in eliminating >99% of skin microfilariae within 1-2months after treatment (Osei-Atweneboana et al., 2007; Basáñez et al., 2008) and in three months all microfilariae in uteri of female worms are expected to have degenerated (Osei-Atweneboana et al., 2011; Nana-Djeunga et al., 2014). However, at standard dose, ivermectin has no measurable effect on oogenesis/embryogenesis or on the amount of sperm and oocytes produced by adult male and female worms (Chavasse et al., 1993). Therefore, the development of oocytes into morulae, coiled and horse-shoe shaped embryos may not be affected by the standard dose of the drug (Schulz-Key and Karam, 1986; Schulz-Key, 1990).

Nonetheless, multiple doses of ivermectin interrupt oogenesis/embryogenesis by lowering the proportion of the female seminal receptacles that contain sperm. This reduces fertilization of oocytes and results in fewer embryonic stages within the uterus (Chavasse et al., 1993).

Due to the continuous usage of ivermectin in endemic regions for many years, some O. volvulus populations have been reported to be responding sub-optimally to treatment (Awadzi et al., 2004a; Awadzi et al., 2004b; Osei-Atweneboana et al., 2007; Osei-Atweneboana et al.,

2011; Pion et al., 2013). After treatment, normal female worm populations may repopulate the skin at about 10% of their baseline microfilarial load within 6 months and 20% within 12 months, but some worm populations have been reported to show much higher repopulation rates (Awadzi et al., 2004a; Awadzi et al., 2004b; Basáñez et al., 2008; Frempong et al.,

2016; Nana-Djeunga et al., 2014; Osei-Atweneboana et al., 2011) [also see Chapter 4]. For such populations, the female worms are capable of producing normal, stretched microfilariae even three months after treatment. Worms which have been exposed to the drug for more

University of Ghana http://ugspace.ug.edu.gh

years have been found to produce oocytes at faster rates compared to their counterparts with no treatments (Nana-Djeunga et al., 2014). This may indicate that repeated or prolonged treatments are directly associated with sub-optimal responses in some parasite populations

(Awadzi et al., 2004a; Awadzi et al., 2004b; Churcher et al., 2009). It is important to understand how these high repopulation rates are correlated with the reproductive status of adult female worms after treatments with standard doses of ivermectin.

The purpose of this chapter is to assess the embryostatic effect of standard doses of ivermectin

(through embryogram analyses) on the adult female worms three months after treatment and investigate any associations between the female reproductive status and the high microfilarial repopulation rates observed (Frempong et al., 2016) in Chapter 4 within some endemic communities with long-term history of treatment in Ghana.

5.2 Methods

5.2.1 Study sites and design

Samples were taken from selected individuals who had been assessed for palpable nodules at pre-treatment in July 2010 as described in Chapter 4. These individuals were selected from 10 communities (in the Brong-Ahafo and Northern Regions, Figure 3.1) that had gone through

15–24 years of annual ivermectin treatment with reports of sub-optimal responses (Awadzi et al., 2004a; Awadzi et al., 2004b; Osei-Atweneboana et al., 2007; Basáñez et al., 2008). In

July 2010, all of these communities were enrolled onto the biannual treatment strategy with mass ivermectin treatment given twice a year. Participants selected for this study had at least one palpable nodule and were successfully treated by direct observation with the standard dose of ivermectin (150 µg/kg body weight) with 3 rounds during the biannual treatment in 76

University of Ghana http://ugspace.ug.edu.gh

July 2010, January 2011 and July 2011 (Figure 3.2). Three months after the last round

(November 2011), nodulectomies were performed to excise nodules for assessment of female reproductive status based on presence of stretched/fully developed microfilariae. Each participant had received at least 10 years of annual treatment before the biannual strategy.

Those selected for nodulectomy consented through written documentation (see Appendix I) after detailed explanation of the processes and documentation of their treatment history.

5.2.2 Nodulectomy

Examination of onchocercal nodules at pre-treatment was done as previously described

(Albiez et al., 1988). Subcutaneous nodules were sought by visual inspection and body palpation. Locations of nodules were recorded on a human anatomical diagram.

Nodulectomies were performed by a trained surgeon under aseptic conditions using local anaesthesia. Area around the nodule was injected with Lignocaine drug to numb the skin. All palpable nodules present (see Figure 5.1) were removed from the participants. Nodulectomies took place at the nearest local health centre to the study community. The excised nodules were placed individually in a Petri dish containing sterile medium 199 solution supplemented with Earle’s salt, L-glutamine (GIBCOTM, Life Technologies, Invitrogen Corporation, USA) and excess tissues removed. All nodules were observed under an inverted microscope for presence of microfilariae around nodules (this was to give an indication if microfilariae had migrated from the uterus of the adult female worm). Nodules were transferred into 50 ml falcon tubes and stored in liquid nitrogen for transportation to the laboratory.

University of Ghana http://ugspace.ug.edu.gh

5.2.3 Nodule digestion, isolation and examination of adult worms

More extraneous tissues were removed from nodules after thawing. The nodules were weighed, measured (length and breadth) and digested using the collagenase technique

(Schulz-Key et al., 1977; Schulz-Key et al., 1980; Schulz-Key, 1988) with slight modifications. Each nodule was placed in a fresh 50 ml tube containing 5–10 ml of 0.5% collagenase A (Roche Diagnostics, Germany) in sterile medium 199 solution supplemented with Earle’s salt, L-glutamine (GIBCOTM, Life Technologies, Invitrogen Corporation, USA),

0.22mg/ml sodium bicarbonate and 0.2 mg/ml gentamicin sulphate. Samples were incubated on a water bath at 37oC and shaking at 150 rpm for 10–20 hours depending upon the weight and size of the nodule. After digestion, worms were isolated gently with the aid of an inverted microscope, a pair of forceps and sterile plastic inoculating loop. Males and females that were found intertwined (Figure 5.2A) were carefully separated with the aid of the pair of forceps.

Worms were rinsed with 0.9% sterile saline and individually placed in Petri dishes. The viability and age of the worms at the time of the nodulectomies were estimated. Worms were classified as alive (viable) prior to nodulectomy if their body architecture was intact/well preserved or if they were moving after their isolation from the nodule. Live worms (at nodulectomy) were also classified based on the condition of their uterine musculature. Broken or damaged worms were not examined for embryograms. The ages of the adult female worms were estimated based on body colour, their size and the prominence of cuticular ridges

(Chavasse et al., 1992; Klager et al., 1996; Specht et al., 2009). The adult female worms that were generally small and transparent were characterized as young, opaque and yellowish as middle-aged and large and brownish worms as old (Schulz-Key, 1988; Specht et al., 2009).

The young female worms were estimated to have lived in the host for less than 5 years while

University of Ghana http://ugspace.ug.edu.gh

middle-aged individuals for 5-7 years and old aged worms for more than 7 years (Specht et al., 2009). The adult female worms that were intact or viable were prepared for embryograms.

5.2.4 Embryogram procedure

The adult female worms selected for embryograms were the only samples used for this study.

Each worm was placed in a glass mortar and cut into pieces with a sterile scalpel. Five hundred microlitres of sterile medium 199 solution was added by rinsing the scalpel into the mortar. A glass pestle was used to homogenize the worm gently but firmly to crush the adult female worm and expel the embryonic stages into the solution without damaging them. An additional 500 µl was used to rinse the pestle to obtain a total of 1 ml homogenate. Ten microlitres of the homogenate was loaded into a Rosenthal counting chamber and the stretched/fully developed intra-uterine microfilariae observed under an inverted microscope and counted with the help of a tally counter as described previously (Schulz-Key et al., 1977;

Schulz-Key, 1988). Normal stretched microfilariae were categorized as smooth, undamaged microfilariae without vacuoles (Figure 5.4A). These were recorded to be alive at the time of nodulectomies. Those with vacuoles (Figure 5.4B) were scored as stretched microfilariae at early degenerating stages and those found to be beaded were treated as late degenerating stretched microfilariae(Figure 5.4C). Degenerated microfilariae were recorded as dead at the time of nodulectomies.

University of Ghana http://ugspace.ug.edu.gh

5.2.5 Data Analyses

Data were analysed using “R” (version 3.2.3) (R Core Team, 2015) and IBM SPSS Statistics

Software Version 20. The average intra-uterine microfilarial counts within the communities were expressed as arithmetic means. The difference in proportions of female worm groups was estimated using Chi2 test. The mean difference in intra-uterine microfilariae was calculated using an independent-sample t-test. The probability of observing normal or degenerated microfilariae in the adult female worms (outcome) in the 10 communities was estimated using a generalized linear mixed model (mixed effects logistic regression) implemented in R (R Core Team, 2015). The model accounted for clustering in the data arising from worms being sampled from the same nodule and nodules from the same patient.

Takumdo (the community with least years of annual treatment i.e. 15 years) was used as the reference community, which also represented the ‘intercept’ of the model. The fitted model output was expressed as the log odds of normal microfilariae (given a total number of normal and degenerated microfilariae). The level for statistical significance for all of the tests was set at P < 0.05.

University of Ghana http://ugspace.ug.edu.gh

5.3 Results

A total of 225 nodules was excised from 106 participants with an average of about 2 nodules per individual (1 - 6 nodules obtained per individual). A total of 495 worms was isolated from the nodules, out of which 67% (N = 331) were female and 33% (N = 164) male, with about 2 female worms and 1 male worm per nodule (Table 5.1). Among the communities,

Jagbengbendo exhibited the highest number of worms (N = 165), representing 33% of the total number of worms isolated/extracted, while Wiae recorded the lowest (9 worms, 1.8% of the total). There was only one volunteer from Wiae whose nodules were removed. The number of female and male worms followed a similar pattern, with Jagbengbendo recording the highest number (127/331 females, 38% and 38/164 males, 23%) while Wiae recorded the least (5/331 females, 1.5% and 4/164 males, 2.4%) out of the total female and total male worms. The overall average number of female worms per nodule was 1.5 (± 0.45) and the average ranged from 0.5 in Ohiampe to 1.8 in Agborlekame 1 and Takumdo (Table 5.1). The overall average number of male worms per nodule was 0.7 (± 0.46) and the average ranged from 0.3 in Ohiampe to 1.8 in Agborlekame 1 (Table 5.1). Out of the 225 nodules digested, about 10% (N = 23) were found to be calcified. All the communities except Asubende and

New-Longoro showed at least one calcified nodule. Ohiampe, Takumdo and Wiae showed

>15% of calcified nodules. The results on nodules and worms from the 10 communities are summarized in Table 5.1.

University of Ghana http://ugspace.ug.edu.gh

Table 5.1: Summary results on nodules and worms obtained from participants Community Years of No. of No. of Total no. Total no. Total no. Mean no. of Mean no. of Mean no. of Mean annual partici- nodules of worms of female of male female male worms female number of ivermectin pants digested (%) worms worms worms per per nodule worms per male worms treatment examined (% (%) (%) nodule (SD) (SD) patient (SD) per patient before 2010 calcified) (SD) Agborlekame 1 24 10 17 (5.8) 62 (12.5) 33 (10.0) 29 (17.7) 1.83 (1.09) 1.81 (1.58) 3.30 (2.21) 2.90 (2.33) Asubende 24 8 13 (0) 29 (5.9) 21 (6.3) 8 (4.9) 1.61 (0.82) 0.72 (0.70) 2.63 (2.67) 1.00 (1.07) Baaya 23 10 13 (7.7) 13 (2.6) 7 (2.1) 6 (3.7) 0.78 (0.67) 0.67 (1.12) 0.78 (0.67) 0.67 (1.12) Jagbengbendo 20 30 76 (9.2) 165 (33.3) 127 (38.4) 38 (23.2) 1.65 (0.87) 0.66 (0.78) 4.00 (2.57) 1.29 (1.16) Kyingakrom 23 7 19 (5.2) 50 (10.1) 31 (9.4) 19 (11.6) 1.51 (0.91) 0.91 (1.02) 4.43 (4.89) 2.71 (2.81) New-Longoro 23 10 20 (0) 40 (8.1) 24 (7.3) 16 (9.8) 1.24 (0.36) 1.24 (0.91) 2.67 (2.18) 1.78 (0.83) Ohiampe 23 7 13 (38.5) 13 (2.6) 8 (2.4) 5 (3.0) 0.5 (0.5) 0.29 (0.49) 1.14 (1.46) 0.71 (1.50) Senyase 23 10 13 (7.7) 22 (4.4) 16 (4.8) 6 (3.7) 1.19 (0.81) 0.42 (0.50) 1.78 (2.11) 0.67 (1.00) Takumdo 15 13 38 (15.8) 92 (18.6) 59 (17.8) 33 (20.1) 1.83 (1.04) 0.94 (0.84) 4.62 (2.87) 2.54 (2.22) Wiae 18 1 3 (33.3) 9 (1.8) 5 (1.5) 4 (2.4) 1.67 (0) 1.33 (0) 5.00 (0) 4.00 (0) Total 495 331 164 Figures in parentheses show the percentage of nodules calcified out of the total number of nodules obtained from each community (column 4). For other percentage estimations (columns 5, 6 and 7), the figures in parentheses are percentages of worms out of the total number of worms obtained from all communities combined. Treatment information given is the number of annual rounds of ivermectin (years) received by each community before 2010, when biannual treatment began. Each individual participant received three rounds of biannual treatment before nodules were excised. SD = standard deviation.

University of Ghana http://ugspace.ug.edu.gh

Figure 5.1: Nodules excised from an onchocerciasis patient

Figure 5.2: Adult female and male worms A: Adult female worm intertwined with male worm. B: The anterior and posterior part of the adult female worm.

University of Ghana http://ugspace.ug.edu.gh

Figure 5.3: Adult male worm (O. volvulus) showing the coiled head region

The reproductive status of female worms was assessed through the embryograms and this was used to determine the response profile of the female worms. A total of 273 out of 331 female worms was analysed by embryogram. Out of this number 37% (N = 273) of female worms had all intra-uterine stretched microfilariae degenerated (dead at the time of nodulectomies,

Figures 5.4B and 5.4C) while only 3% had all intra-uterine stretched microfilariae classified as normal/viable (alive at the time of nodulectomies, Figure 5.4A). Those with both normal and degenerated formed 25% while 35% did not have any intra-uterine stretched microfilariae. Pooling these proportions for females based on viability of microfilariae in the uterus, there was a significantly higher number of female worms (72%) without normal stretched microfilariae (i.e. those with only degenerated and no stretched microfilariae) compared to 28%, which were harbouring normal microfilariae in their uteri at the time of nodulectomies (Chi2 = 53.6; degrees of freedom (d.f.) = 1; P < 0.0001). Other younger embryonic stages such as morulae and coiled forms were observed in some of the samples

(Figure 5.5), but the embryostatic impact of ivermectin was assessed using only stretched microfilariae. Results showed that in all 10 communities, there were female worms

University of Ghana http://ugspace.ug.edu.gh

harbouring normal/viable stretched microfilariae three months after treatment when nodulectomies were performed (Figure 5.6). At Baaya, New-Longoro, Ohiampe and Wiae there was no female worm with only normal microfilariae observed in the uterus (Figure 5.6).

C Figure 5.4: Different forms of stretched microfilariae found in the uteri of adult female worms A: Normal stretched microfilaria (alive at the time of nodulectomy) B: Degenerating stretched microfilaria showing vacuoles (dead at the time of nudulectomy) C: Degenerating stretched microfilaria (beaded form) [dead at the time of nodulectomy]

University of Ghana http://ugspace.ug.edu.gh

Figure 5.5: Embryonic forms of microfilariae found in the uterus of an adult female worm

The proportion of female worms found with only degenerated intra-uterine stretched microfilariae were higher than those with only normal forms in all the communities (Table

5.2). Likewise, the mean number of degenerated intra-uterine stretched microfilariae was higher than the mean number of normal forms in 9 out of the 10 communities (apart from

Asubende), though at Agborlekame 1, Kyingakrom, Ohiampe and Takumdo these differences were not statistically significant (Table 5.2). At Asubende the mean number of normal intra- uterine stretched microfilariae was statistically significantly higher than the mean number of the degenerated forms (P = 0.022). The reproductive status of the female worms from the 10 communities is summarized in Table 5.2. There was a great deal of variation in the mean number of microfilariae from all communities as shown by the value of the standard deviation.

The distribution of the mean number of intra-uterine stretched microfilariae among the communities is also shown in Figure 5.7. Senyase and Wiae showed a high mean number of the microfilariae per female worm (102,600 and 131,400 microfilariae respectively, Figure 86

University of Ghana http://ugspace.ug.edu.gh

5.7). Although this was the case, >85% of the microfilariae were found to be degenerated.

Baaya, Jagbengbendo, Ohiampe and New-Longoro also showed very high numbers of degenerated microfilariae (>75%). At Agborlekame 1, Kyingrakrom and Takumdo the female worms showed about 40% normal and 60% degenerated microfilariae in their uterus (Figure

5.7).

The adult female worms were categorized into three age groups as young, middle-aged and old. About 60% (N = 495) of all the female worms were categorized as middle-aged and the young and old worms were 24% and 16% respectively. A non-parametric chi2 test showed that the proportion of the middle-aged worms was significantly higher than the young and old

(P< 0.0001). In estimating proportions based on reproductive status, the groups of worms classified as young and old had a much higher number of worms with no stretched microfilariae as well as with only normal microfilariae compared with the middle-aged group

(Figure 5.9). Assessment based on mean numbers showed that, the mean number of normal microfilariae in uterus of all the young female worms was 18.7 + 63.7 while the mean number of degenerated microfilariae was 42.0 + 93 (N = 65). In the middle-aged group, the mean number of normal microfilariae was 19.2 + 62.5 and the mean number of degenerated microfilariae was 42.0 + 69 (N= 164). The old worms presented a mean number of normal microfilariae of 10.8 + 32.5 and a mean number of degenerated microfilariae of 45.0 + 94.5

(N = 44). There was no statistically significant difference between the mean number of normal or of degenerated microfilariae among any of the worm age groups.

University of Ghana http://ugspace.ug.edu.gh

Table 5.2: Summary results on the reproductive status of female worms examined through embryogram Community No. of female No. of No. of No. of No. of Mean no. of Mean no. of worms females with females with females with females with normal mf degenerated mf analysed by only normal only normal & no stretched (SD) x 1000 (SD) x 1000 embryogram mf (%) degenerated degenerated mf (%) mf (%) mf (%) Agborlekame 1 33 1 (3.0) 17 (51.5) 7 (21.2) 8 (24.2) 16.1 (61.1) 27.0 (32.1) Asubende 19 1 (5.3) 7 (36.8) 6 (31.6) 5 (26.3) 39.1 (108.9)* 8.4 (11.2)* Baaya 7 0 (0) 2 (28.6) 1 (14.3) 4 (57.1) 0.6 (1.5)* 13.6 (22.8)* Jagbengbendo 85 2 (2.4) 25 (29.4) 25 (29.4) 33 (38.8) 13.5 (33.2)* 41.5 (77.5)* Kyingakrom 31 2 (6.5) 11 (35.5) 8 (25.8) 10 (32.3) 24.3 (70.5) 39.4 (61.0) New-Longoro 24 0 (0) 15 (62.5) 1 (4.2) 8 (33.3) 0.3 (1.4)* 58.8 (76.6)* Ohiampe 8 0 (0) 3 (37.5) 1 (12.5) 4 (50.0) 6.0 (17.0) 21.1 (24.8) Senyase 15 1 (6.7) 3 (20.0) 3 (20.0) 8 (53.3) 14.2 (30.9)* 102.6 (166.3)* Takumdo 46 2 (4.3) 15 (32.6) 15 (32.6) 14 (30.4) 30.0 (84.8) 41.5 (66.4) Wiae 5 0 (0) 3 (60.0) 1 (20.0) 1 (20.0) 3.2 (7.2)* 131.4 (223.3)* Total 273 9 101 68 95 Percentages given in the above table are of female worms out of the total worms embryogrammed in each community. Mean number of intra- uterine stretched microfilariae (mf) were estimated using arithmetic mean. SD = standard deviation. The intra-uterine mf were estimated in a volume of 10µl. The total number of intra-uterine mf in the table are multiplied by 1000. In 90% of the communities the mean intra-uterine mf degenerated was higher than the mean normal mf apart from at Asubende where the mean normal mf was higher than the mean degenerated mf and this was also significant (P = 0.022). Comparing columns 7 and 8: Figures with * show a statistically significant difference between mean number of intra-uterine normal mf and mean number of intra-uterine degenerated mf in each community (P = 0.034 – 0.0001). A total of 196 (101 +95) out of 273 female worms (72%) did not have normal stretched microfilariae in uterus

University of Ghana http://ugspace.ug.edu.gh

Figure 5.6: Percentage distribution of adult female worms with their reproductive status in the 10 communities. All 10 communities showed some female worms producing normal/viable microfilariae (mf) at the time of nodulectomies, though some also had degenerated forms in their uterus. Six out of the 10 communities showed some female worms producing only normal mf.

Figure 5.7: Mean number of degenerate and normal intra-uterine microfilariae per female worm in the 10 study communities. The secondary axis shows the number of adult female worms embryogrammed from the various communities and represented by the green triangles. Averages were calculated using arithmetic means.

University of Ghana http://ugspace.ug.edu.gh

Figure 5.8: Frequency distribution of the reproductive status of female worms by their age. About 60% of female worms were categorized as middle-aged, 16% old and 24% young. All the different reproductive statuses were found in the age groups. The numbers of females producing only normal microfilariae (mf) were similar in all three age groups.

Figure 5.9: Percentage distribution of the reproductive status of female worms by their age. All the different reproductive statuses were found in all the age groups. The proportion of adult female worms producing only normal microfilariae (mf) was highest in old age, followed by the young and then the middle-aged. The proportion with no stretched mf was lowest among the middle-aged group.

University of Ghana http://ugspace.ug.edu.gh

5.3.1 Regression analysis of annual treatment on normal microfilariae production

Data on the number of normal and degenerated stretched microfilariae observed in uterus of female worms were used for the logistic regression analysis. This included information on

291 adult female worms extracted from 182 nodules obtained from 96 patients. There was clustering in the probability of normal microfilariae among individual patients and strong clustering among nodules from the same patient; the random effects variance terms were 2.3 and 20.8 respectively. Generally, the probability or proportion of normal microfilariae (out of the total number of normal and degenerated microfilariae) observed in the uterus of the adult female worms did not vary among communities (Table 5.3). The exception was New-

Longoro, where the coefficient was negative (-8.13) and statistically significantly different from zero (P = 0.0005). This indicates that in this community, the probability of observing normal microfilariae was much less than observing normal microfilariae in the reference community of Takumdo (odds ratio = 3x10-4). Plotting these coefficients (odds ratio) against the number of years of annual treatment before the start of biannual treatment for each community reveals no obvious associations (Figure 5.10).

University of Ghana http://ugspace.ug.edu.gh

Table 5.3: Summary output of generalized linear mixed model (effect of community on normal microfilariae production) Predictor Coefficient in log Coefficient in odds Standard P value variable odds ratio (CI) ratio (CI) error Agborlekame 1 -2.74 (-7.07, 1.59) 0.06 (8.5x10-4, 4.91) 2.21 0.213 Asubende 0.92 (-3.49, 5.33) 2.52 (3.1x10-2, 206.44) 2.25 0.682 Baaya -4.30 (-10.89, 2.29) 0.01 (1.9x10-5, 9.83) 3.36 0.200 Jagbengbendo -1.80 (-5.09, 1.49) 0.16 (6.1x10-3, 4.45) 1.68 0.282 Kyingakrom 0.90 (-3.26, 5.06) 2.46 (3.9x10-2, 156.84) 2.12 0.671 New-Longoro -8.13 (-12.74, -3.52) 3x10-4 (2.9x10-6, 2.9x10-2) 2.35 0.0005* Ohiampe -5.14 (-11.14, 0.86) 5.9x10-3 (1.5x10-5, 2.36) 3.06 0.093 Senyase -2.26 (-8.51, 3.99) 0.10 (2.0x10-4, 54.18) 3.19 0.480 Takumdo # -2.48 (-5.22, 0.26) 1.00 (5.4x10-3, 1.30) 1.40 0.076 Wiae -1.27 (-11.09, 8.55) 0.28 (1.5x10-5, 5164.69) 5.01 0.800 # Takumdo was used as the reference community (predictor variable) which also represented the intercept of the model. Since model outcome is binary, normal microfilariae (mf) represented the success and degenerated mf represented the failure. The odds represent the probability of normal mf per total mf (normal and degenerated) or the proportion of normal mf in total. The outcome of the model shown in the second column is the log odds ratio (odds of normal mf in community/odds of normal mf in Takumdo) whiles the reference community Takumdo is in log odds. Hence a positive log odds ratio indicates a higher probability of observing normal mf in female worms compared to those from Takumdo and vice versa for a negative log odds. Figures in parenthesis represent the confidence interval (CI) and * P value < 0.05 indicates a coefficient that is statistically significantly different from zero.

University of Ghana http://ugspace.ug.edu.gh

Figure 5.10: Relationship between the odds ratio of obtaining normal microfilariae and years of annual treatment in the 10 communities. The y axis represent estimates of odds ratio (odds of observing normal mf in community/odds of observing normal mf in Takumdo) using the generalized linear mixed model and x axis represent years of annual treatment prior to biannual strategy. The horizontal dotted line represent the odd ratio of the reference community, Takumdo (odds ratio = 1) and the bullets represent the estimated odds ratio for each community with reference to Takundo. The vertical lines indicate the 95% confidence interval around the estimates. The confidence intervals were truncated at 20.

University of Ghana http://ugspace.ug.edu.gh

5.4 Discussion

Ivermectin has been used in the control of onchocerciasis in Ghana since 1987, when the first community trials of five rounds of annual ivermectin distribution were conducted in the hyperendemic focus of Asubende (Alley et al., 1994). This treatment strategy has been successful in reducing prevalence and intensity but there is residual transmission in some communities within some endemic regions (Lamberton et al., 2014; Lamberton et al., 2015).

The effect of ivermectin on O. volvulus is twofold, comprising a microfilaricidal effect and an embryostatic effect on the adult female worm. There is evidence that the microfilaricidal effect is still potent in clearing >99% of microfilaridemias one month after treatment (Osei-

Atweneboana et al., 2007; Basáñez et al., 2008). Yet, little is known about the current embryostatic effect on the adult female worm after long-term treatment in endemic communities. Following reports on sub-optimal responses (Awadzi et al., 2004a; Awadzi et al., 2004b; Osei-Atweneboana et al., 2007; Basáñez et al., 2008; Osei-Atweneboana et al.,

2011; Pion et al., 2013) and observation of high repopulation rates in a recent study

(Frempong et al., 2016), this research was carried out to investigate the embryostatic effects of ivermectin on the adult female worm (preventing the release of stretched/live microfilariae from the uterus) three months after treatment during the early stages of a biannual strategy.

Discussions in this chapter will also highlight any associations with the recent observation of high repopulation rates in some of the study communities. Due to the extensive treatment in most of these communities, this study could not make use of an ivermectin-naïve population with which to compare the observations made. The findings here are compared with other similar studies that made use of naïve populations including those that worked in same communities as this study (Awadzi et al., 2004a; Awadzi et al., 2004b; Osei-Atweneboana et al., 2011; Nana-Djeunga et al., 2014).

University of Ghana http://ugspace.ug.edu.gh

Increased frequency of ivermectin treatment has a major effect on nodule number and size, morphology and number of intra-nodular female worms as well as their fertility (Duke et al.,

1990; Chavasse et al., 1992; Klager et al., 1993; Walker et al., 2017). Therefore, the more a population is treated, the more it results in a reduction in the worm population. It was observed in this study that the average number of worms per nodule was much reduced (1.4 females and 0.7 males) compared to 2-2.3 females and 1.3-1.5 males for untreated/naïve worm populations in a similar study (Nana-Djeunga et al., 2014) and even for a study that worked in similar communities in the past (Osei-Atweneboana et al., 2011). Yet, comparing results with the treated group (annual single dose) in these two studies, there was no obvious difference in the worm numbers per nodule. Although Gardon et al. (2002) showed that repeated treatments at standard or higher doses are associated with increased mortality in the adult female worm population, it is unclear whether the increased frequency of treatment

(biannual) in this study made any significant impact in reducing worm numbers per nodule.

Since only three rounds of biannual treatments were given before nodulectomies were performed, it is possible that the effect of the biannual strategy in reducing worm numbers may be observed some years later. Considering the fact that some of these communities have a history of sub-optimal responses to ivermectin (Awadzi et al., 1997; Awadzi et al., 2004a;

Osei-Atweneboana et al., 2007; Churcher et al., 2009; Osei-Atweneboana et al., 2011) with high rates of skin repopulation by microfilariae (Frempong et al., 2016), the biannual treatment strategy is still necessary with additional interventions to continue reducing infection levels if elimination of the disease is to be achieved.

Worm populations that are untreated and those that respond poorly to treatment have been observed to have high proportions of nodules containing viable/normal worms and

University of Ghana http://ugspace.ug.edu.gh

microfilariae (Osei-Atweneboana et al., 2011). Nana-Djeunga et al. (2014) also showed that about 32% of female worms within the control group with no history of large-scale ivermectin treatment were found with all intra-uterine stretched microfilariae degenerated (i.e. nodulectomies that were performed before treatment with ivermectin). This shows that even in the absence of treatment, a proportion of female worms at any time may have microfilariae unable to leave the uterus which then degenerate. It was, therefore, expected that the percentage of adult female worms with all intra-uterine stretched microfilariae degenerated would be higher than 37% (as observed in this study), considering the fact that three rounds of biannual treatment were given before nodulectomies were performed and that these were carried out three months after the last round. All the same, about 35% of the females were observed without any stretched/matured microfilariae in their uteri (although some had other embryonic stages). Therefore, the impact of biannual treatment cannot be overlooked; at the time nodulectomies were performed, a significant proportion of the female worms (about

72%, P < 0.0001) were not producing viable/normal microfilariae and would not have been able to contribute to transmission (i.e. release microfilariae that would migrate to the skin).

Although it has been shown that the biannual treatment made an impact in suppressing female reproductive activity, there are still concerns that in all 10 communities some female worms were observed to be producing normal microfilariae three months after treatment, when the embryostatic effect should still be apparent (Figure 5.6 and Table 5.2). Previous studies in these communities have shown an average repopulation rate of about 50% of baseline loads within 6 months post treatment in almost all communities (Frempong et al., 2016) which agrees with this observation. These high repopulation rates have been associated with long

University of Ghana http://ugspace.ug.edu.gh

exposure of parasite population to ivermectin treatment in other studies (Awadzi et al., 2004a;

Awadzi et al., 2004b; Churcher et al., 2009; Nana-Djeunga et al., 2014).

The embryostatic effect of ivermectin treatment is associated with more degenerated microfilariae observed in female uteri as well as more calcified nodules/worms observed in treated populations (Awadzi et al., 2004b; Osei-Atweneboana et al., 2011; Nana-Djeunga et al., 2014). Therefore, nodules excised from good responding communities are expected to have more female worms harbouring degenerated microfilariae and/or high calcification of nodules/worms. This was the case at Baaya, Ohiampe, Wiae (good responding by Osei-

Atweneboana et al., 2007 and Frempong et al., 2016) and New-Longoro. In these communities all the female worms harbouring normal microfilariae were also found with degenerated forms and none was observed to have only normal microfilariae as observed at

Agborlekame 1, Asubende, Jagbengbendo, Kyingakrom, Senyase and Takumdo (Figure 5.6).

Furthermore in the good responding communities, there was a much higher frequency of calcified nodules (>15%) compared to the other communities (0 - 9%, Table 5.1). Some of the communities such as Jagbengbendo, Senyase and Takumdo, although classified as good responding in a previous study (Frempong et al., 2016) showed some female worms with only normal microfilariae in the uteri (poorly responding).

The mean number of intra-uterine stretched microfilariae estimated in each community was not directly proportional to the number of female worms sampled in these communities

(Figure 5.7). In Jagbengbendo, the highest number of female worms were embryogrammed, yet Senyase and Wiae, where fewer female worms were analysed showed the highest mean number of intra-uterine microfilariae with 102,600 and 131,400 respectively (Figure 5.7 and

University of Ghana http://ugspace.ug.edu.gh

Table 5.2). The net microfilarial productivity in a population may not necessarily depend on the total number of adult female worms, particularly if there is density-dependent fecundity

(Duerr et al., 2004). Also, not all female worms may be producing microfilariae at the same time or there may be few females producing most of the microfilariae (Hildebrandt et al.,

2012; Hildebrandt et al., 2014). In Table 5.2 almost all communities showed higher mean numbers of degenerated microfilariae than mean number of normal forms. Asubende, one of the communities with the highest number of years of previous treatment and categorized as poorly responding (Osei-Atweneboana et al., 2007) with significantly high microfilarial repopulation rates (Frempong et al., 2016) was the only community with a significantly higher mean number of normal microfilariae compared to the degenerated forms. This study did not show any associations between the number of years of prior ivermectin treatment and skin microfilarial repopulation rates (Frempong et al., 2016; Chapter 4) as well as years of treatment and reproductive status of female worms (this Chapter, Figure 5.10). Though this is the case, Asubende and Kyingakrom have consistently shown reduced effects of ivermectin over the years. There is the need for a more intensive intervention (perhaps quarterly treatments) or alternative treatment strategies (anti-wolbachial therapies and/or vector control) to reduce transmission in these communities and the other communities where high levels of normal microfilarial production were observed (Agborlekame 1, Jagbengbendo, Senyase and

Takundo) including communities previously described as good responding (Osei-

Atweneboana et al., 2007; Frempong et al., 2016). These responses have been attributed to individual female worms responding poorly to treatment and driving the response profile of the female population in these communities (Osei-Atweneboana et al., 2011; Pion et al.,

2013; Frempong et al., 2016). Individually, unique capabilities of poor responses have also been associated with genetic factors (Kudzi et al., 2010; Osei-Atweneboana et al., 2012).

University of Ghana http://ugspace.ug.edu.gh

The age of female worms categorized in this study was based only on morphological keys

(Schulz-Key, 1988) which has some limitations. Other studies have considered ageing worms based on morphological and physiological differences which are more reliable (Specht et al.,

2009). Additionally, some communities recorded few worms due to low numbers of participants willing to be nodulectomized. Therefore the age distribution reported here may not give a true reflection of the entire female worm population. Nonetheless this study found a younger female population (16% old, 60% middle-aged and 24% young) compared with Osei-

Atweneboana et al., (2011) [65% old, 26% middle-aged and 9% young] who employed similar age categorization and worked in some of these communities. Age estimations of worms have been used to assess the intensity of transmission in communities where it was not possible to estimate the annual transmission potential (Duke et al., 2002). Therefore the young worm population observed here is attributed to continuous transmission over the years with an introduction of new worms into the population.

In Figure 5.8, all age groups recorded some female worms with normal and degenerate microfilariae and those without microfilariae in uteri at the time of the nodulectomies.

Assessing distribution based on the proportions (Figure 5.9), there were more old-aged worms producing normal microfilariae than middle-aged or young ones. This agrees with observations that older worms are more likely to respond sub-optimally due to longer exposure to ivermectin treatment (Osei-Atweneboana et al., 2011). Likewise the proportion of female worms without intra-uterine microfilariae was also high among the old and young worms (Figure 5.9). Reasons may be that some of these young worms were immature and had not started producing oocytes at the time of nodulectomies. Additionally, some older female

University of Ghana http://ugspace.ug.edu.gh

worms are normally found with degenerated or empty uteri due to a reduced ability to produce oocytes (Schulz-Key and Karam, 1986; Specht et al., 2009).

This study has shown that ivermectin is still capable of preventing the release of stretched intra-uterine microfilariae in a high proportion of female worms, but indicates that some individual worms continue to produce normal microfilariae after treatment. This was observed in all the 10 communities studied, without a clear association with the number of years of treatments received. A step-up of treatment coverage and compliance may be needed, or the frequency of treatment may have to be increased, or alternative treatment strategies may have to be introduced in order to reach the 2020 / 2025 elimination targets set by the World Health

Organization (WHO, 2012a; WHO, 2012b).

100

University of Ghana http://ugspace.ug.edu.gh

CHAPTER SIX

SINGLE NUCLEOTIDE POLYMORPHISMS WITHIN BETA-TUBULIN AND P-GLYCOPROTEIN GENES ASSOCIATED WITH ONCHOCERCA VOLVULUS SUB-OPTIMAL RESPONSES TO IVERMECTIN TREATMENT

6.1 Introduction

Although anticipated and feared, it is not yet clear whether the widespread use of ivermectin for treating onchocerciasis during the second half of the Onchocerciasis Control Programme

(OCP) in West Africa (Alley et al., 1994; Boatin, 2008) and in subsequent years, has exerted a sizeable and measurable selection pressure on the genome of Onchocerca volvulus. The principle of selection is that, as the drug is continuously used, some parasite traits may be favoured that increase fitness in the presence of selection (Abel, 2009). These traits would have a genetic basis, and hence genetic changes associated with responses to the drug (in this case ivermectin) would indicate the occurrence of genetic structuring in O. volvulus populations. Such treatment-induced selection would be heritable and subsequent generations that will acquire this trait may survive the effect of the drug.

The sub-optimal/poor response to ivermectin treatment observed in O. volvulus populations

(Awadzi et al., 2004a; Awadzi et al., 2004b; Osei-Atweneboana et al., 2007; Pion et al.,

2013) is of considerable concern to onchocerciasis control programmes. This is because ivermectin is currently the only drug used for mass treatment of the disease. The drug is still potent in clearing >99% of microfilariae in the skin of hosts and temporarily prevents the release of intra-uterine stretched microfilariae from adult female worms, as previously discussed in Chapter 5. Poorly responding females are deemed to recover from the

101

University of Ghana http://ugspace.ug.edu.gh

embryostatic effects earlier and repopulate the skin with new microfilariae at rates faster than expected. The exact cause of this response is not yet known. Some studies have reported the existence of poorly responding worm populations and others have shown some genetic basis to this response (Awadzi et al., 2004a; Awadzi et al., 2004b; Osei-Atweneboana et al., 2007;

Osei-Atweneboana et al., 2012; Pion et al., 2013). Yet, others have reported that the poor response observed in worm populations may not have a genetic basis but associated with programmatic factors such as low treatment coverage and compliance (Burnham, 2007; Cupp et al., 2007; Mackenzie, 2007; Remme et al., 2007). The nature of the response of O. volvulus populations to ivermectin is controversial because the extent of the natural variation between individual host responses in treatment-naïve communities is not well understood (Churcher et al., 2009; Awadzi et al., 2014) and there are few comprehensive studies that have monitored treatment in individual hosts, categorized worm responses to ivermectin treatment and supported findings with genetic evidence (Bourguinat et al., 2007; Osei-Atweneboana et al.,

2012). Most of such studies have been cross-sectional, observing differences in allele frequencies within treated and untreated groups but not showing single nucleotide polymorphisms (SNPs) associated with poor response phenotype (Ardelli and Prichard, 2004;

Ardelli et al., 2005; Eng et al., 2006; Ardelli et al., 2006b; Bourguinat et al., 2008).

Eng and Prichard (2005) analysed the polymorphism of some resistant candidate genes in

O. volvulus. These genes were considered candidate genes because of their association with ivermectin resistance in parasitic nematodes of farmed ruminants, and included: P- glycoprotein (P-gp), glutamate-gated chloride channel, P-glycoprotein-like proteins, beta- tubulin (β-tub) isotype 1 and gamma-aminobutyric acid (GABA). Among these genes, P-gp and β-tub showed a statistically significant difference in genetic polymorphism between

102

University of Ghana http://ugspace.ug.edu.gh

ivermectin-treated and untreated worms in the study. As much as these reports were helpful and created awareness about possible resistance development in O. volvulus populations, they did not relate the polymorphisms to response phenotypes (good or poor) within the population but only to allele frequencies in treated and untreated groups. The responses of parasites in relation to their genetic changes are relevant in determining whether treatment with ivermectin in a population will fail or not, irrespective of changes in allele frequencies.

Nonetheless, some studies have indeed related genetic changes in parasite populations to their responses to ivermectin treatment (Bourguinat et al., 2007; Kudzi et al., 2010; Osei-

Atweneboana et al., 2012). Bourguinat et al. (2007) showed that treatment with ivermectin over time reduced the frequency of the β-tub “aa” homozygotes from 68.6% to 25.6%, while the “ab” heterozygotes increased from 20.9% to 69.2% in female worm samples. These changes in allele frequencies were associated with reduced fertility within the homozygote group and increased fertility within the heterozygote group. A more recent study showed that changes in the allele frequencies of O. volvulus multidrug resistance genes and cytochrome

P450 were associated with poor responses in onchocerciasis patients exposed to ivermectin for a long time (Kudzi et al., 2010). Osei-Atweneboana et al. (2012) also showed that genetic polymorphism within the β-tub gene is associated with poor responses to ivermectin treatment in O. volvulus populations. In the latter study, 8 out of 24 single nucleotide polymorphisms

(SNPs) assessed were found to occur at significantly higher (P < 0.05) frequencies in sub- optimal/poor response communities compared with good responding communities as well as ivermectin-naïve communities. The phenotypic and genotypic analyses showed that the genotype (1183GG/1188CC/1308TT/1545GG) was strongly associated with sub-optimal response to ivermectin treatment (Osei-Atweneboana et al., 2012). One of the SNP positions

103

University of Ghana http://ugspace.ug.edu.gh

(1545) was also observed to occur at different frequencies among ivermectin-treated and untreated worms in another study in Cameroon (Nana-Djeunga et al., 2012), confirming some genetic relationship to sub-optimal responses in O. volvulus populations.

The purpose of this chapter is to explore regions within the β-tub and P-gp genes where SNPs have been previously shown to be associated with ivermectin treatment groups, identify them and investigate the association between these SNPs and ivermectin response phenotypes. This study is a two-year follow up of biannual ivermectin treatment where individual host responses were monitored through skin snipping. Categorization of worm response to treatment considered both responses in hosts and in individual female worms. The outcome of this research will provide preliminary data on the regions within β-tub and P-gp genes that may need further research to develop a molecular marker for identifying poorly responding worms.

104

University of Ghana http://ugspace.ug.edu.gh

6.2 Methods

This study was carried out in 10 communities in the Brong-Ahafo and Northern regions of

Ghana (Figure 3.1; also refer to Chapter 3 for details). A total of 956 participants was randomly selected from different households at the beginning of the study and tested by skin snipping (July 2010). The first round of biannual treatment was done by direct observation during this survey. Six months later (January 2011), the participants were again skin snipped and treated by direct observation during the second round of biannual treatment (both microfilaria-positive and negative individuals at pre-treatment). A total of 217 individuals who were positive for microfilariae and palpable nodules was then followed up after January

2011 in April 2011, July 2011 and November 2011. Individuals who provided written consent were nodulectomized in November 2011, i.e. three months after the third round of biannual treatments (July 2011). The nodules were digested using the collagenase technique and embryogram analyses were performed (Schulz-Key et al., 1977; Schulz-Key et al., 1980;

Schulz-Key, 1988) to assess the reproductive status of female worms at the time of the nodulectomies (refer to Chapters 4 and 5 for procedures). Based on the phenotypic response to treatment by skin snipping (host) and the reproductive status of the female parasite (worms) from the embryogram data, the adult female worms were grouped into good and poor responders to ivermectin treatment as described below (section 6.2.1).

6.2.1 Characterization of adult female worm responses (selection of worms)

Adult female worm response to ivermectin treatment was characterized based on host response in skin and the reproductive status of the female worm. Individuals who were positive for skin microfilariae in April 2011 and/or November 2011, i.e. three months after treatment were considered as poor responding hosts. It is expected that the lowest

105

University of Ghana http://ugspace.ug.edu.gh

microfilaridermia levels (near zero) are reached 1-2 months after treatment, and these levels should be maintained beyond three months after treatment (Basáñez et al., 2008). Individuals who did not show any skin microfilaria in April 2011 and/or November 2011 were considered as good responding hosts. Adult female worms extracted from good responding hosts were separated from poor responding hosts. The worms from the poor responding hosts with all or

>80% of intra-uterine stretched microfilariae stages described as normal i.e. stretched microfilariae alive at the time of nodulectomies (see Figure 5.4A), were classified as poor responding worms. Worms classified as good responders were those female worms obtained from good responding hosts with all the intra-uterine stretched microfilariae degenerated

(dead, Figures 5.4B and 5.4C), or which contained no stretched microfilariae. There were however, few female worms obtained from hosts classified as poor responders, but harboured almost equal numbers of degenerated and normal intra-uterine stretched microfilariae. These were classified as intermediate responders. A total of 60 female worms (30 good responders,

20 poor responders and 10 intermediate responders) was identified to fit the classifications of ivermectin response and selected for beta-tubulin (β-tub) amplification. Forty (20 good responders, 15 poor and 5 intermediate) out of the 60 female worms were selected for P-gp amplification in 6 different regions of the gene. The worms were selected from Agborlekame

1, Asubende, Kyingakrom and Ohiampe (Kintampo/Pru Districts) in the Brong-Ahafo Region as well as Jagbengbendo and Takumdo (Kpandai District) in the Northern Region.

6.2.2 DNA extraction from adult female worms

DNA was extracted from the selected good, intermediate and poor responding adult female worms. The extreme ends of the posterior or anterior part of the female worms (the portion that the uterus does not extend to) were used for the extractions. This was to prevent the

106

University of Ghana http://ugspace.ug.edu.gh

possibility of DNA “contamination” from sperm, fertilized eggs or microfilariae within the uteri of female worms. DNA was extracted using the BIOLINE ISOLATE II, Genomic DNA kit-Protocol (Bioline, 2016) with slight modifications as follows:

1) A 2 cm length of the posterior/anterior part of female worms was placed into a 1.5 ml

micro-centrifuge tube and incubated at -80oC overnight.

2) Tissue was immediately grounded into pieces using a sterile plastic pestle.

3) 180 µl lysis buffer (GL) was used to wash the pestle into the tube and 25 µl of

Proteinase K was added to the solution and vortexed.

4) The sample was incubated overnight at 56oC in a shaking incubator (150 rpm).

5) 200 µl of another lysis buffer (G3) was added to the sample, vortexed vigorously and

incubated at 70°C for 10 minutes.

6) 210 µl of absolute ethanol (96-100%) was added to the sample and vortexed

vigorously.

7) The solution from 6) was transferred into the ISOLATE II Genomic Spin Column and

placed on a collection tube.

8) The sample was spun for 1 minute at 11,000 x g and the flow-through discarded.

9) DNA retained on the membrane was washed by adding 500 µl of wash buffer (GW1)

and centrifuged for 1 minute at 11,000 x g. The flow-through was discarded and the

collection tube re-used.

10) -A volume of 600 µl of wash buffer (GW2) was added to the column and centrifuged

for 1 minute at 11,000 x g. The flow-through was discarded and the collection tube re-

used.

107

University of Ghana http://ugspace.ug.edu.gh

11) The sample was centrifuged for 1 minute at 11,000 x g, to remove residual ethanol and

the ISOLATE II in the Genomic Spin Column placed on a fresh and sterile 1.5 ml

micro-centrifuge tube.

12) DNA was eluted into the tube by adding 100 µl of pre-heated elution buffer (70°C)

directly onto the membrane, incubated for 5 minutes and centrifuged at 11,000 x g for

1 minute.

6.2.3 Estimation of DNA concentration

The concentration of DNA was estimated with 1 µl of each extracted sample using the Qubit assay kit and measured with the Qubit® 2.0 fluorometer (Invitrogen™) following the manufacturer’s protocol.

6.2.4 DNA amplification

DNA was amplified using the polymerase chain reaction (PCR) with specific primers designed for one region within the β-tub gene (Osei-Atweneboana et al., 2012) and 6 regions within the P-gp gene (Ardelli et al., 2005; Ardelli et al., 2006a). The P-gp regions included:

Trans membrane domain 1: i) Iodide Transport defect 1 (TMD1-ITD1); ii) Iodide Transport defect 4 (TMD1-ITD4); iii) Trans membrane 5 (TMD1-TM5); iv) Nucleotide binding domain

1: Walker 1 (NBD1-WA1); v) Nucleotide binding domain 2: Iodide Transport defect 14

(NBD2-ITD14), and vi) the Linker regions. The primer pair information is given in Table 6.1.

108

University of Ghana http://ugspace.ug.edu.gh

PCR amplification was done in a 35 µl reaction containing 3 µl DNA template (4 – 5 ng), 1X

PCR buffer, 1.5 mM MgCl2, 0.4 mM DNTPs mix, 0.2 µM forward primer, 0.2 µM reverse primer and 1 unit Taq polymerase enzyme. The cycling conditions used for amplification were as follows:

Initial denaturing of 94oC for 2 minutes, followed by 33 cycles of (94oC for 30 seconds, 50-

55oC for 45 seconds & 72oC for 1 minute). Then a final extension of 72oC for 5 minutes.

109

University of Ghana http://ugspace.ug.edu.gh

Table 6.1: Primer sequences and expected diagnostic band sizes for amplification Gene Primer name Primer Sequence Genomic Annealing Band position of temperature size primer (oC) (bp) P-gp TMD1-ITD1-F1 5-TGACAGAGCTGAAAACTAATG-3 25 - 804 50 780 TMD1-ITD1-R1 5-TAATTCCGATTGATGCATAAAG-3

P-gp TMD1-ITD4-F1 5-CATTCAAGTAGCATGCAGAGC-3 1615 - 1809 53 195 TMD1-ITD4-R1 5-GAAGAGACCATCGAAAAAACC-3

P-gp TMD1-TM5-F1 5-CATGTCGTAAATTTGCATTGC-3 1751 - 2265 50 515 TMD1-TM5-R1 5-TCCCGGATTCATTATTATACG-3

P-gp NBD1-WA1-F1 5-GTGAGATAGAATTTCAAAACG-3 3074 - 4052 50 979 NBD1-WA1-R1 5-ATGTTAATTCCATCAATCAGG-3

P-gp NBD2-ITD14-F1 5-ATTATGGTTTCTGGCTGTTTG-3 8585 - 9100 55 516 NBD2-ITD14-R1 5-ACTTCCAGATGGTCCAGTGAC-3

P-gp LINKER-FI 5-CATCATTGGTCAAGTCACAGC-3 5271 - 5673 50 403 LINKER-R1 5- CCACTCATAATACTACCTCGT-3

β-tub IVM_B-tub-F1 5-CGTCTGGCATTGAGTATTAG-3 1002 - 1687 50 706 IVM_B-tub-R1 5-CCACATCTGAACTTAAAATGC-3 The region of interest within beta-tubulin and p-glycoprotein have also been described by Eng and Prichard 2005, Ardelli et al., 2005 & 2006; Osei-Atweneboana et al., 2012. Primers were designed specifically to amplify areas within these regions of interest.

110

University of Ghana http://ugspace.ug.edu.gh

6.2.5 Gel electrophoresis

Five microlitres of the PCR products were run on a 1.5 – 2% agarose gel (depending on size of fragment) stained with 0.13 µg/ml ethidium bromide for 1 hour. These were visualized using a UV trans-illuminator (BioDoC-It™ Imaging System, Benchtop UV Trans- illuminator) and the images saved. 30 µl of PCR products were prepared for sequencing at

McGill University, Montreal, Quebec, Canada.

6.2.6 Data analysis

Sequence results were analysed using Sequencher 5.0 software (multiple alignment; https://www.genecodes.com) to identify SNPs associated with the responses to ivermectin treatment (refer to Appendix VI for chromatograms of DNA sequences). The reference β-tub and P-gp sequences were obtained from the National Center for Biotechnology Information

(NCBI) GenBank: AF019886.1 and AY884214.1, respectively. Pair-wise comparisons of the genotype and allele frequencies of worms from the three responding groups (good, intermediate and poor) were carried out to determine the differences in frequencies within β- tub and P-gp genes using Chi2 tests. Observed genotype and allele frequencies were compared with expected frequencies estimated using the Hardy-Weinberg equation (HWE). The local inbreeding coefficient (F-statistic) of worms sampled from Kintampo/Pru and Kpandai

Districts was estimated for genetic diversity in these two populations. The reduction in observed heterozygosity was compared with expected heterozygosity estimated using HWE

(Appendix VII A, B and C). All statistical analyses were estimated using 95% confidence interval and significance set at P < 0.05 (α = 0.05).

111

University of Ghana http://ugspace.ug.edu.gh

6.3 Results

A multiple alignment with the reference sequences for β-tub (GenBank: AF019886.1) and P- gp (GenBank: AY884214.1) showed 10 SNP positions which were polymorphic. Five of the

10 SNPs were identified within the β-tub gene at positions 1183 T/G, 1188 T/C, 1298 T/C,

1308 C/T and 1545 A/G. The other 5 SNPs were identified within the P-gp gene at positions

1744 C/T, 2002 T/G, 5403 T/C, 5505 C/T and 5546 A/G. The SNPs within the P-gp were found within the TMD1-ITD4, TMD1-TM5 and the Linker regions. Three SNPs out of the 10 identified were significantly associated with a poor response phenotype at positions 1308 and

1545 within the β-tub gene (Pearson chi2 = 8.3; d.f. = 2; P = 0.016 and Pearson chi2 = 9.7; d.f.

= 2; P = 0.008) and at position 5546 within the P-gp gene (Pearson chi2 = 7.6; d.f. = 2; P =

0.023). There was a complete nucleotide change from A to G at position 1555 within the β- tub gene which occurred in all of the samples (position fixed). Assessing the allele frequencies among the three response groups (estimated from Tables 6.2 and 6.3), there was a significantly higher nucleotide change among the poor responding worms compared to the good responders at positions 1545 within β-tub (Pearson chi2 = 8.2; d.f. = 1; P = 0.004) and

5546 within P-gp (Pearson chi2 = 6.3; d.f. = 1; P = 0.012). Allele frequency estimation formula is shown in Appendix VII A.

112

University of Ghana http://ugspace.ug.edu.gh

A B Figure 6.1: Gel electrophoresis pictures showing DNA fragments Panel A: Lane 1 = 100bp DNA ladder, lanes 2, 3 & 6 = TMD1-ITD1 (780bp), lanes 4&5 = IVM_B- tub (706bp), lanes 7-10 = TMD1-ITD4 (195bp), lanes 11-15 = NBD2-ITD14 (516bp), lane 16 = Negative control. Panel B: Lane 1 = 100bp DNA ladder, lanes 2-5 = TMD1-ITD1 (780bp), lanes 6-10 = LINKER (403bp), lanes 11-13 = TMD1-TM5 (515bp), lane 14 = unsuccessful amplification, lanes 15-16 = NBD1-WA1 (979bp), lane 17 = Negative control

113

University of Ghana http://ugspace.ug.edu.gh

A combined analysis of the genotype frequencies showed that the heterozygote genotype

CT/AG was dominant in all the three response groups but in different proportions. About 70% of poor responding worms were found with the genotype CT/AG; this genotype was present in 50% of the intermediate and 35% of good responders. The genotype CT/AA was also present in all three ivermectin response groups but in equal proportions (about 9% in each group). The homozygote genotype combination of CC/AA, similar to the reference sequence from GenBank, was found in about 50% of the good responders and 20% in the intermediate responders but absent in the poor responding worms. The genotypes CT/GG and TT/GG were only present in the intermediate and poor responding worms but absent in the good responding worms (Figure 6.2).

Due to the low number of worms from each community, the genetic diversity of each worm population was estimated at the district level between Kintampo/Pru and Kpandai using the local inbreeding coefficient with information from the genotype frequencies in Table 6.4. The estimated values for the inbreeding coefficient for Kintampo/Pru and Kpandai are given in

Table 6.5. The heterozygosity within the β-tub gene at positions 1308 and 1545 for

Kintampo/Pru was found to be reduced compared to the expected heterozygosity. At these same positions, the population from Kpandai was found to have gained heterozygosity the observed frequency being higher than expected (i.e. outbreeding). Nonetheless, both districts were observed to have gained more heterozygosity than expected within the P-gp gene at position 5546. After estimation of Hardy-Weinberg Equilibrium (HWE) at the 10 SNP positions identified within β-tub and P-gp genes, only position 1308 within the β-tub gene was statistically significantly different from the HWE (P = 0.03).

114

University of Ghana http://ugspace.ug.edu.gh

Table 6.2: Genotype frequencies at five SNP positions for beta-tubulin SNP at genomic sequence positions for beta-tubulin (β-tub) isotype 1 gene Ivermectin No. of 1183 (TT) 1188 (TT) 1298 (TT) 1308 (CC)* 1545 (AA)* response worms TT TG GG TT TC CC TT TC CC CC CT TT AA AG GG n(%) n(%) n(%) n(%) n(%) n(%) n(%) n(%) n(%) n(%) n(%) n(%) n(%) n(%) n(%) Good 24 21 3 0 (0) 8 12 4 24 0 (0) 0 (0) 11 11 2 15 8 1 (87.5) (12.5) (33.3) (50) (16.7) (100) (45.8) (45.8) (8.3) (62.5) (33.3) (4.1)

Intermediate 10 7 (70) 2 (20) 1 2 (20) 8 0 (0) 8 (80) 2 0 (0) 2 (20) 7 (70) 1 3 (30) 5 (50) 2 (20) (10) (80) (20) (10)

Poor 12 8 4 0 (0) 1 9 2 11 1 0 (0) 0 (0) 11 1 1 9 (75) 2 (66.7) (33.3) (8.3) (75) (16.7) (91.7) (8.3) (91.7) (8.3) (8.3) (16.7)

N = total number of worms successfully sequenced; n = number of a particular genotype at an SNP position; figures in parentheses are the percentages of total number of worms successfully sequenced. *shows a statistically significant difference in the genotype frequencies at that SNP position between the good and poor response groups (at position 1308, P = 0.016 and position 1545, P = 0.008).

115

University of Ghana http://ugspace.ug.edu.gh

Table 6.3: Genotype frequencies at five SNP positions within P-glycoprotein gene SNP at genomic sequence positions for P-glycoprotein (P-gp) gene Response 1744 (CC) 2002 (TT) 5403 (TT) 5505 (CC) 5546 (AA)* CC CT TT TT TG GG TT TC CC CC CT TT AA AG GG n(%) n(%) n(%) N n(%) n(%) n(%) N n(%) n(%) n(%) N n(%) n(%) n(%) N n(%) n(%) n(%) N Good 8 3 0 (0) 11 7 (70) 3 (30) 0 (0) 10 10 4 0 (0) 14 10 4 0 (0) 14 9 5 0 (0) 14 (72.7) (27.3) (71.4) (28.6) (71.4) (28.6) (64.3) (35.7)

Intermediate 1 2 0 (0) 3 1 2 0 (0) 3 2 1 0 (0) 3 2 1 0 (0) 3 1 1 1 3 (33.3) (66.7) (33.3) (66.7) (66.7) (33.3) 66.7) (33.3) (33.3) (33.3) (33.3)

Poor 6 8 0 (0) 14 8 5 0 (0) 13 10 3 (23) 0 (0) 13 10 3 (23) 0 (0) 13 2 9 2 13 (42.9) (57.1) (61.5) (38.5) (77) (77) (15.4) (69.2) (15.4)

N = total number of worms successfully sequenced. n = number of a particular genotype at an SNP position; figures in parentheses are the percentages of total number of worms successfully sequenced. *shows a statistically significant difference in the genotypic frequencies at that SNP position between the good and poor response groups (at position 5546, P = 0.023).

116

University of Ghana http://ugspace.ug.edu.gh

The heterozygote genotype CT/AG was dominant in all the three response groups but in different proportions. About 70% of poor responding worms were found with the genotype

CT/AG; this genotype was present in 50% of the intermediate and 35% of good responders

(Figure 6.2). The genotype CT/AA was also present in all three ivermectin response groups but in equal proportions (about 9% in each group). The homozygote genotype combination of

CC/AA, similar to the reference sequence from GenBank, was found in about 50% of the good responders and 20% in the intermediate responders but absent in the poor responding worms. The genotypes CT/GG and TT/GG were only present in the intermediate and poor responding worms but absent in the good responding worms (Figure 6.2).

Due to the low number of worms from each community, the genetic diversity of each worm population was estimated at the district level between Kintampo/Pru and Kpandai using the local inbreeding coefficient with information from the genotype frequencies. The estimated values for the inbreeding coefficient for Kintampo/Pru and Kpandai are given in Table 6.5.

The heterozygosity within the β-tub gene at positions 1308 and 1545 for Kintampo/Pru was found to be reduced (inbreeding) compared to the expected heterozygosity. At these same positions, the population from Kpandai was found to have gained heterozygosity (observed frequency higher than expected, i.e. outbreeding). Nonetheless, both districts were observed to have gained more heterozygosity than expected within the P-gp gene at position 5546

(Table 6.5). After estimation of Hardy-Weinberg Equilibrium (HWE) at the 10 SNP positions identified within β-tub and P-gp genes, only position 1308 within the β-tub gene was statistically significantly different from the HWE (P = 0.03, Table 6.6).

117

University of Ghana http://ugspace.ug.edu.gh

Table 6.4: Genotype frequencies within Kintampo/Pru and Kpandai districts for three SNP positions Genotypic frequencies at the district level for SNP positions 1308, 1545 and 5546 District 1308 (CC) 1545 (AA) 5546 (AA) CC CT TT AA AG GG AA AG GG n(%) n(%) n(%) N n(%) n(%) n(%) N n(%) n(%) n(%) N Kintampo/Pru 4 (33.3) 5 (41.7) 3 (25) 12 4 (33.3) 5 (41.7) 3 (25) 12 7 (33.3) 11 (52.4) 3 (14.3) 21

Kpandai 9 (26.5) 24 (70.6) 1 (2.9) 34 15 (44.1) 17 (50) 2 (5.9) 34 5 (55.6) 4 (44.4) 0 (0) 9

N = total number of worms successfully sequenced. n = number of a particular genotype at an SNP position; figures in parentheses are the percentages of total number of worms successfully sequenced. Allele frequencies were estimated from this table to calculate the local inbreeding coefficient (F-statistic) in Table 6.5 (refer to Appendix VII).

118

University of Ghana http://ugspace.ug.edu.gh

Table 6.5: Estimated values of inbreeding coefficient (F-statistic) for Kintampo/Pru and Kpandai districts District SNP Observed Observed Expected Expected Inbreeding position homozygosity heterozygosity homozygosity heterozygosity coefficient (frequency) (frequency) (HWE) (HWE) Kintampo/Pru 1308 0.583 0.417 0.503 0.497 0.161

1545 0.583 0.417 0.503 0.497 0.161

5546 0.476 0.524 0.518 0.482 -0.087

Kpandai 1308 0.294 0.706 0.528 0.472 -0.495

1545 0.500 0.500 0.573 0.427 -0.171

5546 0.556 0.444 0.654 0.346 -0.286

The inbreeding coefficient is given by FIS = (Expected Heterozygosity-Observed Heterozygosity)/Expected Heterozygosity. The observed homozygosities and heterozygosities were estimated from the genotype frequencies in Table 6.4 (see Appendix VII for estimations). The expected homozygosities and heterozygosities were estimated using the Hardy-Weinberg equation (HWE) with information from allele frequencies estimated from Table 6.4 (see Appendix VII for estimations). Positive inbreeding coefficient indicate a loss of heterozygosity (in-breeding) due to non-random mating within sub-populations and negative values indicate a gain in heterozygosity (out-breeding).

119

University of Ghana http://ugspace.ug.edu.gh

Table 6.6: Hardy-Weinberg Equilibrium at 10 SNP positions identified within beta-tubulin and p-glycoprotein genes Gene Region SNP Position HWE (Chi2 P-value value)

β-tub β-tub 1183 0.23 0.632

β-tub β-tub 1188 3.50 0.06

β-tub β-tub 1298 0.05 0.82

β-tub β-tub 1308 4.45 0.03*

β-tub β-tub 1545 0.45 0.50

P-gp TMD1-ITD4 1744 2.56 0.11

P-gp TMD1-TM5 2002 1.47 0.23

P-gp Linker 5403 0.71 0.40

P-gp Linker 5505 0.71 0.40

P-gp Linker 5546 0.29 0.59

The Hardy-Weinberg Equilibrium (HWE) in the above table were estimated using the genotype frequencies given in Tables 6.2 and 6.3. Refer to Appendix VII for the estimation of HWE. *Indicates that the genotypic frequencies at that SNP position (observed) are statistically significantly different from the HWE (expected).

120

University of Ghana http://ugspace.ug.edu.gh

Figure 6.2: Combined genotype frequencies within the three ivermectin responding groups at β-tubulin positions 1308 and 1545. Some genotypes were found in poor responding worms (CT/GG and TT/GG), while others were found in good responder worms (CC/AA and TT/AA). Intermediate responders shared some genotypes with both good and poor responders.

121

University of Ghana http://ugspace.ug.edu.gh

6.4 Discussion

This chapter characterized the responses in treated hosts as well as worm populations following detailed two-year follow-ups after treatments in some onchocerciasis endemic regions in Ghana. These responses have been supported with preliminary genetic data to show changes that are associated with poor and good response phenotypes to ivermectin treatment.

It is hoped that the results in this research will address some of the controversies surrounding sub-optimal/poor response to ivermectin treatment, and indicate further research avenues to be explored for identification of possible markers of ivermectin poor response.

The genes studied here have already been associated with changes in allele frequencies within ivermectin-treated populations (Eng and Prichard, 2005; Eng et al., 2006; Ardelli et al.,

2006a; Ardelli et al., 2006b). The challenge with relating genetic changes in treated and untreated groups is the lack of evidence to show that these changes are reflected in ivermectin response phenotypes. Genetic changes within the worm population (indicating genetic structuring because of geography, genetic drift, selection, etc.) may not be very relevant to control programmes, instead how any of these changes could relate programmatically to treatment failure or poor response to treatment.

This research was unable to make use of untreated groups (ivermectin-naïve) with which to compare observations. This was due to the extensive treatment schedules that have taken place within endemic communities in Ghana over the years, starting with the initial community trials of ivermectin during the Onchocerciasis Control Programme (OCP) era, and continuing with the devolution of onchocerciasis control activities after the closure of the

OCP. Endemic communities had received between 15-24 years of annual ivermectin

122

University of Ghana http://ugspace.ug.edu.gh

treatment in the past and all the participants selected here had received at least 10 years of annual treatment before the start of this study. Nonetheless, the aim of exploring single nucleotide polymorphisms (SNPs) associated with sub-optimal/poor responses to treatment has been, in part, achieved. It must be said that this study was not aimed to be a genome-wide study but focused on candidate genes that had already been investigated.

All the 10 SNP positions: β-tub (1183T/G; 1188T/C; 1298T/C; 1308C/T; 1545A/G) and P-gp

(1744C/T; 2002T/G; 5403T/C; 5505C/T, 5546A/G) were found to be polymorphic and analysed in this study have also been described in other studies as SNPs in O. volvulus worms obtained in areas under ivermectin treatment (Ardelli et al., 2006a; Nana-Djeunga et al.,

2012; Osei-Atweneboana et al., 2012). Osei-Atweneboana et al. (2012) showed that changes in four positions (1183T/G; 1188T/C; 1308C/T, and 1545A/G) were strongly associated with poor ivermectin response. Out of these mutations, the 1308C/T and 1545A/G were the ones associated with poor response to ivermectin (P = 0.016 and 0.008, respectively) based on the phenotypic response classification adopted here in this study. The agreement between these two studies about the identification of these two SNP positions is an indication that a poor response phenotype may indeed be associated with genetic changes within the β-tub gene.

Likewise, changes in some loci within the P-gp gene (ITD1; ITD4; TM5, and ITD14) have been shown to occur at different frequencies within ivermectin-treated and untreated groups of worms (Ardelli et al., 2005). In that study there was no difference in allele frequencies within the Linker region for the two groups (Ardelli et al., 2005). In contrast, based on phenotypic response, the Linker region was found to be polymorphic in this study, and changes at position 5546A/G were statistically significantly associated with poor ivermectin responses (Table 6.3). It is possible that the low number of worms sequenced might have

123

University of Ghana http://ugspace.ug.edu.gh

affected the power to observe differences at more SNP positions within the P-gp gene. A better approach would have been a complete sequencing of the 10,574bp length of P-gp and

3696bp of β-tub to effectively identify all SNPs relating to poor response using high- throughput sequencing as suggested by (Kotze et al., 2014). Due to constraints with logistics, the study focused on regions that had been previously described to show changes in allele frequencies among treated and untreated worms (Ardelli et al., 2006a; Nana-Djeunga et al.,

2012; Osei-Atweneboana et al., 2012). All the same, there is some evidence that at least one

SNP position within P-gp and two positions within β-tub may be associated with a poor response phenotype to ivermectin treatment as classified here.

It is also worth noting that the classification employed here may have some limitations since poor and good responses were based on positive or negative microfilariae in skin 3 months post treatment. Classifications based on skin microfilariae are very difficult to accurately categorize since even in ivermectin-naïve communities where it is expected to have good responses, there are still inter-individual variation in responses to treatment (Churcher et al.,

2009; Awadzi et al., 2014). Therefore to keep classification simple, this study settled on host with positive and negative skin microfilariae 3 months post treatment as the cut-off for poor and good responses, respectively. The negative microfilariae in the skin 3 months post treatment may not mean true absence of microfilaridermia due to the low sensitivity of the skin snips technique (Bottomley et al., 2016). Nonetheless the strength in this classification is borrowed from the information on the adult female worms (embryogram data) as an additional criterion for selection of worms. In Chapters 4 and 5, it was observed that generally repopulation rates were high in all communities (Figure 4.4) and this is reflected in the embryogram showing all communities with females harbouring normal microfilariae (Figure

124

University of Ghana http://ugspace.ug.edu.gh

5.6). Furthermore Asubende and Kyingakrom with statistically significantly high repopulation rates compared to the reference community, Takumdo (Figure 4.4) showed a relatively higher mean normal microfilariae than the rest of the communities (Figure 5.7). This information is indicative of some association between embryogram status of worms and responses in host skin within the communities.

The observation of SNPs in β-tub and P-gp associated with poor response is also indicative that if resistance to ivermectin were to develop in O. volvulus populations, this would not be related to a single gene (monogenic), but may involve multiple genes (polygenic) as reported in other helminths (Kotze et al., 2014).

The good and poor response phenotypes were also observed to have distinct genotype combinations (Figure 6.2). The genotypes CT/GG and TT/GG at SNP positions 1308 and

1545 respectively were found in the poor responders but not in the good responding worms.

On the other hand, the homozygous genotype of CC/AA similar to the β-tub reference sequence and genotype TT/AA was found among the good responders but not among the poor responders (Figure 6.2). These genetic differences may be associated with differences in the capability/ability of female worms to produce normal microfilariae even after exposure to the drug (reduced embryostatic effect in poor responders but optimum effect in good responders).

Interestingly the intermediate group shared genetic information with both the good and poor responders which was not surprising (Figure 6.2). The genotype combination also indicates that β-tub gene is indeed associated with a poor response phenotype which agrees with other studies (Bourguinat et al., 2007; Osei-Atweneboana et al., 2012). The genotype combination within the P-gp was not analysed since only one SNP position was observed to be statistically

125

University of Ghana http://ugspace.ug.edu.gh

significantly associated with the poor response phenotype. It can also be inferred from this research that it is more appropriate to classify O. volvulus female worm responses to ivermectin treatment using both host response and embryostatic responses instead of considering only host responses (skin snip data) as reported in some studies (Awadzi et al.,

2004a; Osei-Atweneboana et al., 2011).

Developing a molecular marker to monitor drug efficacy in worm populations will need genetic information as described in this preliminary study. Fortunately, the two SNPs identified within β-tub gene (1308 and 1545) and the additional SNPs shown by Osei-

Atweneboana et al. (2012) at positions 1183 and 1188 all occur within a short length of

362bp. It is, therefore, feasible to develop a molecular marker to identify poor ivermectin response in worm populations without any longitudinal study.

The local inbreeding coefficient (F-statistic) estimated here measures the genetic diversity in a population based on the loss or gain of heterozygosity compared to expected heterozygosity using Hardy-Weinberg equation (HWE) i.e. in-breeding and out-breeding. Analysis of worm populations from Kintampo/Pru district showed a loss in heterozygosity at two SNP positions out of three (Table 6.5). This indicates a reduced introduction of new alleles into the worm population at Kintampo/Pru, a process likely to be encouraged by in-breeding (Shane, 2005).

A possible explanation could be the geographical separation of worm sub-populations within the Kintampo/Pru. In Figure 3.1 (map of study sites), it is observed that in the Brong-Ahafo

Region, communities within the Pru area are geographically distant from those in the

Kintampo area. This might have created a sub-population of locally breeding worms due to distance, encouraging homozygosity and reducing heterozygosity as described by (Shane,

126

University of Ghana http://ugspace.ug.edu.gh

2005). This phenomenon is unlikely to have occurred in Kpandai district since the worm population (communities) is not geographically separated to create a sub-population of locally breeding worms. Kpandai was observed to have gained heterozygosity in all three SNP positions (Table 6.5). In this district most communities exist within the same transmission zone, therefore flies biting in one community may have procured blood meals from other communities (Lamberton et al., 2016) compared to Kintampo/Pru.

There are implications with in-breeding where genetic traits are localized in a particular population without new traits being introduced. If the trait results in a genetic capacity to reduce the effect of the drug on worm populations, resistance development is very rapid.

Nonetheless, if the trait does not support increased ability to fend-off drug effects, elimination of the disease may easily be achieved. In Kintampo/Pru reports show a consistently reduced drug (embryostatic) efficacy over time. In Chapter 4, a number of communities from this district were classified as poor responding (Asubende, Kyingakrom and New-Longoro). On the other hand Wiae, Jagbengbendo and Takumdo (Kpandai district) were identified as having better responses to ivermectin treatment. The Kintampo/Pru district also has a history of long- term treatment (23-24years) compared to Kpandai (15-20years). One would expect more selection of genetic traits to have occurred in the former district, yet, no statistical association of phenotypic response with the number of annual treatments was identified in either Chapter

4 or Chapter 5. Other factors may therefore contribute to this observation other than selection.

The confirmation of sub-optimal/poor response having a genetic basis serves as a warning for the control and elimination of onchocerciasis in Ghana. Currently, a number of countries are considering eliminating this disease with mass treatment of ivermectin (Katabarwa et al.,

127

University of Ghana http://ugspace.ug.edu.gh

2012; Evans et al., 2014). Others have considered stopping MDA after achieving very low prevalence of the disease (Cruz-Ortiz et al., 2012; Traore et al., 2012; Lovato et al., 2014). It is, therefore, important to advance research into genetic evidence of O. volvulus sub- structuring leading to poor responses to ivermectin (due to selection or genetic drift) in order to change treatment strategies intelligently. Frempong et al. (2016) considered that biannual treatment may not assuage concerns of sub-optimal responses to ivermectin, and perhaps just increasing the frequency of treatment (to 6- or 3-monthly) is not enough to curtail transmission. Identification of those individuals responding sub-optimally and harbouring poor response worms, followed by their treatment with other therapies (e.g. anti-wolbachial treatment) should be considered. This therefore makes the development of genetic markers to monitor ivermectin efficacy very essential, by ensuring that areas showing sub-optimal response are identified quickly and the needed interventions implemented.

128

University of Ghana http://ugspace.ug.edu.gh

CHAPTER SEVEN

GENERAL DISCUSSION, CONCLUSION AND RECOMMENDATIONS

7.1 General Discussion

In 2010, the Ghana Health Services (GHS) changed the onchocerciasis treatment strategy from annual to biannual – 6-months treatment (Turner et al., 2013a). This was in response to previous reports of sub-optimal/poor response (faster repopulation of skin with microfilariae) in treated patients (Awadzi et al., 2004a; Awadzi et al., 2004b; Osei-Atweneboana et al.,

2007; Basáñez et al., 2008; Churcher et al., 2009). After decades of mass treatment with ivermectin in Ghana, residual transmission still exists within some hotspots (Lamberton et al.,

2015). However, the actual cause of sub-optimal responses to treatment in O. volvulus populations is unclear. This thesis sought to assess the impact of 3 years of biannual treatment strategy in Ghana (2010 – 2013), determine the embryostatic effect of drug on adult female worm productivity and link any responses in parasite population to genetic changes that have occurred within the population.

In Chapter 4, the 5-6 rounds of biannual treatment strategy (from July 2010 to June 2013) implemented by the Neglected Tropical Disease Control Programme (NTDCP) of the GHS was observed to have made a significant impact in reducing infections. This is shown in the general decline of the community microfilarial prevalence (CMFP) and the community microfilarial loads (CMFLs) after baseline (Figures 4.1 and 4.2 respectively). At the end of the survey in June 2013, the CMFLs had reduced by >40% in most of the communities and the CMFP was significantly <10% in 5 out of 10 communities (see Appendices III and IV).

129

University of Ghana http://ugspace.ug.edu.gh

The impact was also made evident in Chapter 5 where there was a relatively reduced average number of worms per nodule i.e. 1.4 females and 0.7 males per nodule (Table 5.1) after three rounds of biannual treatment compared with a naïve population in a similar study in the past

(Osei-Atweneboana et al., 2011). When worm numbers were compared with that study during annual treatment strategy within these same communities, there was no obvious difference

(Osei-Atweneboana et al., 2011). It is likely that the three rounds of biannual treatment were not enough to show any difference in average worm numbers per nodule between annual and biannual treatment. However, in Chapter 5, a greater proportion of the female worms (about

72%, Table 5.2) were observed without normal stretched microfilariae (found with either degenerated or no stretched microfilariae) at the time nodulectomies were performed. This reduction in female productivity could be attributed to the treatment impact. It is therefore obvious that during the three rounds of biannual treatment, the microfilaricidal and/or embryostatic effect of ivermectin made a significant impact in the worm populations. The macrofilaricidal effect (killing of worms) may probably be observed in the later years of biannual treatments.

Irrespective of the positive treatment impact, there are still concerns in some few communities whether biannual treatment is enough to meet the WHO elimination target by 2020 / 2025

(WHO, 2012a; WHO, 2012b). It is very worrying that in communities like Agborlekame 1,

Asubende, Jagbengbendo and Kyingakrom prevalence was more than 10% at the end of the study in 2013 (Table 4.2), despite many years of mass treatment. Such communities may need to scale-up interventions to meet the WHO set target. Basanez et al. (2008) showed that in a naïve worm population, skin repopulation with microfilariae 6 months after treatment is about

10% of baseline loads (geometric mean). In comparison, the 6-month repopulation after

130

University of Ghana http://ugspace.ug.edu.gh

treatment was very high in some communities reaching up to 50% of pre-treatment levels

(Figure 4.4). These were also higher than those estimated from some of the same communities in a previous study (Osei-Atweneboana et al., 2007), where majority showed <30%. Indeed, some of these discrepancies in the repopulation rates may be due to the fact that here in this study, the repopulation rates were based on arithmetic mean (model output) compared with the 10% geometric mean (Basáñez et al., 2008). Nonetheless, Asubende, Kyngakrom and

New-Longoro with high repopulation rates as shown in Chapter 4 (Figures 4.4 and 4.5A) have also been implicated to respond sub-optimally to treatment based on geometric mean estimations (Osei-Atweneboana et al., 2007) and this should be of great concern to control programmes. Asubende (one of the communities with the highest years of annual treatment before this study) showed a significantly higher mean normal intra-uterine microfilariae than degenerated forms (Table 5.2), though there has not been any obvious associations between the years of annual ivermectin treatment and microfilarial repopulation rates or reproductive status of adult female worms.

The extensive treatment over the years (probably resulting in selection pressure) is likely to have contributed to a reduction in embryostatic effect in worm population. Female worms now recover quickly from treatment effect and begin producing normal microfilariae as early as three months after treatment (Nana-Djeunga et al., 2014). These individual female worms responding sub-optimally to treatment drive the response profile of the entire community by influencing the microfilarial loads (Frempong et al., 2016). Therefore it is not surprising that in all the 10 communities some female worms were observed harboring normal stretched microfilariae i.e. still producing normal microfilariae after treatment at the time of nodulectomies (Figure 5.6). This agrees with the high repopulation rates in hosts skin

131

University of Ghana http://ugspace.ug.edu.gh

observed in almost all 10 communities (Figure 4.4). The sub-optimal response here cannot be attributed to poor treatment coverage as suggested (Cupp et al., 2007; Mackenzie, 2007), since there were high reported treatment coverage achieved throughout the period of study from July 2010 to June 2013 (42 – 93%; Figures 4.1 and 4.2) and participants were treated by direct observation.

The lifespan of Onchocerca volvulus adult worm is approximately 12-15 years and the female worm can produce microfilariae for 9-11 years (Plaisier et al., 1991). As these worms age, their microfilariae productivity reduce (Schulz-Key and Karam, 1986; Specht et al., 2009) which may reduce their contribution to transmission. Therefore the ages of worms in a population are important in onchocerciasis transmission studies. The adult female worms were aged based on morphological keys as described (Schulz-Key, 1988), a method that was used in the past to age worms within most of the communities studied here (Osei-

Atweneboana et al., 2011). Results showed a relatively younger female worm population

(16% old, 60% middle-aged and 24% young) compared with the past study (65% old, 26% middle-aged and 9% young). The presence of a younger worm population may indicate new worms introduced into the population, suggesting continuous transmission in these communities. This is also evident in the microfilariae prevalence observed at the end of study where some communities showed 10-26% CMFP (Table 4.2). Age estimation of worms have also been used to assess the intensity of transmission in communities where it was not possible to estimate the annual transmission potential (Duke et al., 2002). Transmission could also be due to reinvasion of infected blackflies from neighboring endemic communities or countries.

132

University of Ghana http://ugspace.ug.edu.gh

The controversies surrounding sub-optimal responses (Cupp et al., 2007; Mackenzie, 2007;

Osei-Atweneboana et al., 2007) is partly due to lack of follow-up studies to categorize worm responses to ivermectin treatment and evidence to show that genetic changes in worm population are associated with poor response. The true definition of “sub-optimal” response in worm population and the use of appropriate tools for detecting infections (skin snipping not very sensitive at low infections) are also some of the issues to contend with. To address these challenges, this study categorized response profile after treatment using information from both the microfilariae response in host skin (Chapter 4) and embryostatic response of adult female worms (the presence of stretched microfilariae in uterus after treatment, Chapter 5).

The genetic analyses in Chapter 6 (Tables 6.2 and 6.3) showed that all the 10 studied single nucleotide polymorphism (SNP) positions had also been associated with ivermectin treatment in other studies (Ardelli et al., 2006a; Nana-Djeunga et al., 2012; Osei-Atweneboana et al.,

2012). Two of these SNPs within the beta-tubulin gene (β-tub) were statistically significantly associated (P = 0.016 and 0.008, respectively, Table 6.2) with a poor response to ivermectin treatment (female worms producing normal microfilariae 3 months after treatment). This observation agrees with 2 of 4 SNP positions within β-tub strongly associated with a poor response phenotype to ivermectin treatment in a previous study (Osei-Atweneboana et al.,

2012). In addition, within β-tub, there were some specific genotype combinations found in good responders but absent in poor responders, vice versa (Figure 6.2). These genotypes may be associated with differences in the female worms to produce normal microfilariae even after exposure to the drug i.e. reduced embryostatic effect in poor responders but optimum effect of drug in good responders. The genotype combination also indicates that β-tub gene is indeed associated with a poor response phenotype which agrees with other studies (Bourguinat et al.,

2007; Osei-Atweneboana et al., 2012). Genetic sub-structuring resulting in poor response to

133

University of Ghana http://ugspace.ug.edu.gh

treatment due to selection or genetic drift may be occurring in β-tub as shown in the significantly different genotype frequencies from Hardy-Weinberg equation (HWE) at position 1308 (Table 6.6). One SNP position within P-glycoprotein gene at position 5546 A/G associated with a poor response phenotype here (P = 0.023, Table 6.3) but was not found to be significantly higher in an ivermectin treated worm population in another study (Ardelli et al., 2005).

Mating behavior in a worm population affect the genetic divergence of traits from one generation to the other (Churcher et al., 2008). In a non-random mating population, with time a sub-population may be created and traits may be conserved within these sub-population due to inbreeding (Shane, 2005). In Chapter 6, the local inbreeding coefficient (F-statistics) was used as a measure of genetic diversity (Shane, 2005) based on the loss or gain of heterozygosity compared with expected i.e. HWE (see Appendix VII). Analyses of worm populations showed a loss in heterozygosity at two SNP positions out of three in

Kintampo/Pru district (Table 6.5). In this district (within the Brong-Ahafo region), the communities located in the Kintampo area are geographically distant from the Pru area (see map in Figure 3.1) which might have created sub-populations due to distance. Such a situation is likely to encourage in-breeding and reduce transfer of genetic traits among these two sub- populations. In the Kpandai district (within the Northern region), the communities are not geographically distant from each other (Figure 3.1) to create sub-populations among the worms (i.e. communities exist within the same transmission zone). Additionally, mass treatment with ivermectin in Kintampo/Pru district has gone on for many more years than in

Kpandai district (23-24 years and 15-20 years respectively) which might have resulted in

134

University of Ghana http://ugspace.ug.edu.gh

more selection of genetic traits occurring in the former district and reducing genetic diversity in that worm population.

Indeed, there is evidence that O. volvulus sub-optimal response to ivermectin is associated with some level of genetic changes in worm population. These information call for critical monitoring of worm population. Additional interventions needed to supplement mass drug administration (MDA) with ivermectin in Ghana if elimination of onchocerciasis is hoped to be achieved by 2020 / 2025, especially in sub-optimally responding communities.

7.1.1 Strengths and limitations of the study

The major strength of this study is the multiple sampling of a cohort of participants from 10 different communities within two endemic regions in Ghana, to assess the impact of ivermectin treatment on host skin and intra-uterine microfilariae loads. These participants were treated by direct observation and treatment information documented during the study.

Therefore sub-optimal responses to treatment observed in parasite population cannot be attributed to missing of treatment rounds during the study. Additionally, the three response groups (good, intermediate and poor) associated with the genotypes were categorized based on microfilariae responses in hosts’ skin and the embryostatic response of adult female worms. Others have categorized this based on only microfilariae responses in hosts’ skin

(Awadzi et al., 2004a; Osei-Atweneboana et al., 2007; Osei-Atweneboana et al., 2011).

The limitations of this study include the absence of an ivermectin-naïve population (due to the extensive treatment in most of these communities) to directly compare the observations made.

135

University of Ghana http://ugspace.ug.edu.gh

During the genetic studies, some of the sequences were not successful, thus, the numbers analysed were lower than anticipated. Furthermore, for large genes such as β-tub and P-gp, the Sanger sequencing method used could not sequence entire gene at a go, unless multiple primers were used to amplify and sequence the different regions of the entire gene. Due to limited resources, only the regions within the genes known to be associated with ivermectin treated population were selected for amplification and sequencing. These could have contributed to the fewer SNP positions identified to be associated with a poor response phenotype. All the same, this study was able to show that some SNPs within β-tub and P-gp are associated with a poor response phenotype to ivermectin treatment.

7.2 General Conclusion

Data from this research work have shown that the 3 years (5-6 rounds) biannual treatment strategy in Ghana made positive impact in reducing microfilariae loads and adult female worm productivity. Though this is the case, transmission still exists within some areas especially in communities which had been implicated as responding sub-optimally in the past.

The young worm population observed also indicate on-going transmission. Transmission is suggested to be driven by a few sub-optimally responding female worms in each community.

SNPs that have been associated with ivermectin-treated groups in other studies have also been shown here to be associated with a poor response phenotype at positions 1308 (C/T) and 1545

(A/G) within the β-tub gene and position 5546 (A/G) within P-gp gene. Genetic processes leading to sub-structuring were shown to occur within the β-tub gene in the worm populations observed. The adult female worms sampled from Kintampo/Pru district were found to have a reduced heterozygosity (in-breeding) compared to those from the Kpandai district, perhaps

136

University of Ghana http://ugspace.ug.edu.gh

indicating different transmission zones. Constant monitoring of parasite responses to ivermectin treatment is necessary to avoid the emergence of resistant population. Based on these findings, it is uncertain if increasing the frequency of treatment in Ghana will be sufficient to meet the World Health Organization’s 2020 / 2025 elimination goals.

7.3 Recommendations

Future studies on onchocerciasis in Ghana may need to follow-up as many participants as possible in the communities in addition to those positive at baseline. This is to help understand the dynamics of CMFP and CMFL. Estimation of repopulation rates at the individual level instead of community based is also important in the future to help assess the dynamics of repopulation at the individual level. In order to have a stronger statistical power and identify more SNPs, future studies should analyze more samples. Additionally, the whole gene sequence of 10,574bp of P-gp and 3696bp of β-tub should be sequenced using a more advanced sequencing technique such as high-throughput (Kotze et al., 2014). It is also important to conduct a similar study about six to ten years after the start of biannual treatment to investigate a long term impact of biannual treatment, especially in areas where sub-optimal responses have been documented.

137

University of Ghana http://ugspace.ug.edu.gh

REFERENCES

Abel D. L. (2009). The GS (genetic selection) Principle. Front Biosci (Landmark Ed) 14: 2959-2969. Adler P. H., Cheke R. A. and Post R. J. (2010). Evolution, epidemiology, and population genetics of black flies (Diptera: Simuliidae). Infect Genet Evol 10(7): 846-865. Akogun O. B. (1999). Onchocerciasis in Taraba State, Nigeria: Clinical-Epidemiological Study of at-Risk Males in Bakundi District. Zentralblatt für Bakteriologie 289(3). Albers A., Esum M. E., Tendongfor N., Enyong P., Klarmann U., Wanji S., Hoerauf A. and Pfarr K. (2012). Retarded Onchocerca volvulus L1 to L3 larval development in the Simulium damnosum vector after anti-wolbachial treatment of the human host. Parasit Vectors 5: 12. Albiez E. J., Buttner D. W. and Schulz-Key H. (1984). Studies on nodules and adult Onchocerca volvulus during a nodulectomy trial in hyperendemic villages in Liberia and Upper Volta. II. Comparison of the macrofilaria population in adult nodule carriers. Tropenmed Parasitol 35(3): 163-166. Albiez E. J. (1985). Calcification in adult Onchocerca volvulus. Trop Med Parasitol 36(3): 180-181. Albiez E. J., Buttner D. W. and Duke B. O. (1988). Diagnosis and extirpation of nodules in human onchocerciasis. Trop Med Parasitol 39 Suppl 4: 331-346. Aleyasin H., Karuppagounder S. S., Kumar A., Sleiman S., Basso M., Ma T., Siddiq A., Chinta S. J., Brochier C., Langley B., Haskew-Layton R., Bane S. L., Riggins G. J., Gazaryan I., Starkov A. A., Andersen J. K. and Ratan R. R. (2015). Antihelminthic Benzimidazoles Are Novel HIF Activators That Prevent Oxidative Neuronal Death via Binding to Tubulin. Antioxid Redox Signal 22(2): 121-134. Ali M. M., Mukhtar M. M., Baraka O. Z., Homeida M. M., Kheir M. M. and Mackenzie C. D. (2002). Immunocompetence may be important in the effectiveness of Mectizan (ivermectin) in the treatment of human onchocerciasis. Acta Trop 84(1): 49-53. Ali M. M., Baraka O. Z., AbdelRahman S. I., Sulaiman S. M., Williams J. F., Homeida M. M. and Mackenzie C. D. (2003). Immune responses directed against microfilariae correlate with severity of clinical onchodermatitis and treatment history. J Infect Dis 187(4): 714-717. Ali M. M. M. (2006). Immunologic aspects of the pathogenesis of human onchocerciasis Available at: http://www.diva-portal.org/smash/get/diva2:199307/FULLTEXT01.pdf; Accessed on 23/08/2015. Alley E. S., Plaisier A. P., Boatin B. A., Dadzie K. Y., Remme J., Zerbo G. and Samba E. M. (1994). The impact of five years of annual ivermectin treatment on skin microfilarial loads in the onchocerciasis focus of Asubende, Ghana. Trans R Soc Trop Med Hyg 88(5): 581-584. Amazigo U. and Boatin B. (2006). The future of onchocerciasis control in Africa. Lancet 368(346-347). Amazigo U. (2008). The African Programme for Onchocerciasis Control (APOC). Ann Trop Med Parasitol 102 Suppl 1: 19-22. Amin M. L. (2013). P-glycoprotein Inhibition for Optimal Drug Delivery. Drug Target Insights 7: 27-34.

138

University of Ghana http://ugspace.ug.edu.gh

Ardelli B. F. and Prichard R. K. (2004). Identification of variant ABC-transporter genes among Onchocerca volvulus collected from ivermectin-treated and untreated patients in Ghana, West Africa. Ann Trop Med Parasitol 98(4): 371-384. Ardelli B. F., Guerriero S. B. and Prichard R. K. (2005). Genomic organization and effects of ivermectin selection on Onchocerca volvulus P-glycoprotein. Mol Biochem Parasitol 143(1): 58-66. Ardelli B. F., Guerriero S. B. and Prichard R. K. (2006a). Characterization of a half-size ATP-binding cassette transporter gene which may be a useful marker for ivermectin selection in Onchocerca volvulus. Mol Biochem Parasitol 145(1): 94-100. Ardelli B. F., Guerriero S. B. and Prichard R. K. (2006b). Ivermectin imposes selection pressure on P-glycoprotein from Onchocerca volvulus: linkage disequilibrium and genotype diversity. Parasitology 132(Pt 3): 375-386. Ashraf S., Beech R. N., Hancock M. A. and Prichard R. K. (2015). Ivermectin binds to Haemonchus contortus tubulins and promotes stability of microtubules. Int J Parasitol 45(9-10): 647-654. Awadzi K., Opoku N. O., Attah S. K., Addy E. T., Duke B. O., Nyame P. K. and Kshirsagar N. A. (1997). The safety and efficacy of amocarzine in African onchocerciasis and the influence of ivermectin on the clinical and parasitological response to treatment. Ann Trop Med Parasitol 91(3): 281-296. Awadzi K., Boakye D. A., Edwards G., Opoku N. O., Attah S. K., Osei-Atweneboana M. Y., Lazdins-Helds J. K., Ardrey A. E., Addy E. T., Quartey B. T., Ahmed K., Boatin B. A. and Soumbey-Alley E. W. (2004a). An investigation of persistent microfilaridermias despite multiple treatments with ivermectin, in two onchocerciasis- endemic foci in Ghana. Ann Trop Med Parasitol 98(3): 231-249. Awadzi K., Attah S. K., Addy E. T., Opoku N. O., Quartey B. T., Lazdins-Helds J. K., Ahmed K., Boatin B. A., Boakye D. A. and Edwards G. (2004b). Thirty-month follow-up of sub-optimal responders to multiple treatments with ivermectin, in two onchocerciasis-endemic foci in Ghana. Ann Trop Med Parasitol 98(4): 359-370. Awadzi K., Attah S. K., Addy E. T., Opoku N. O., Quartey B. T., Lazdins-Helds J. K., Ahmed K., Boatin B. A., Boakye D. A. and Edwards G. (2004b). Thirty-month follow-up of sub-optimal responders to multiple treatments with ivermectin, in two onchocerciasis-endemic foci in Ghana. Annals of Tropical Medicine & Parasitology 98(4): 359-370. Awadzi K., Opoku N. O., Attah S. K., Lazdins-Helds J. and Kuesel A. C. (2014). A randomized, single-ascending-dose, ivermectin-controlled, double-blind study of moxidectin in Onchocerca volvulus infection. PLoS Negl Trop Dis 8(6): e2953. Basáñez M. G., Pion S. D. S., Churcher T. S., Breitling L. P., Little M. P. and Boussinesq M. (2006). River blindness: a success story under threat? PLoS Medicine 3: 1454- 1460. Basáñez M. G., Pion S. D., Boakes E., Filipe J. A., Churcher T. S. and Boussinesq M. (2008). Effect of single-dose ivermectin on Onchocerca volvulus: a systematic review and meta-analysis. Lancet Infect Dis 8(5): 310-322. Bioline (2016). Purifying genomic DNA from cultured cells and human or animal tissue. Available at: http://www.bioline.com/fr/downloads/dl/file/id/1208/isolate_ii_genomic_dna_kit_prot ocol.pdf; Accessed on 25/08/2016.

139

University of Ghana http://ugspace.ug.edu.gh

Blackhall W. J., Pouliot J.-F., Prichard R. K. and Beech R. N. (1998). Haemonchus contortus: Selection at a Glutamate-Gated Chloride Channel Gene in Ivermectin and Moxidectin Selected Strains. EXPERIMENTAL PARASITOLOGY 90: 42-48. Boakye D. A. (1993). A pictorial guide to the chromosomal identification of members of the Simulium damnosum Theobald complex in west Africa with particular reference to the Onchocerciasis Control Programme Area. Trop Med Parasitol 44(3): 223-244. Boakye D. A., Back C., Fiasorgbor G. K., Sib A. P. and Coulibaly Y. (1998). Sibling species distributions of the Simulium damnosum complex in the west African Onchocerciasis Control Programme area during the decade 1984-93, following intensive larviciding since 1974. Med Vet Entomol 12(4): 345-358. Boatin B. (2008). The Onchocerciasis Control Programme in West Africa (OCP). Ann Trop Med Parasitol 102 Suppl 1: 13-17. Bottomley C., Isham V., Vivas-Martinez S., Kuesel A. C., Attah S. K., Opoku N. O., Lustigman S., Walker M. and Basanez M. G. (2016). Modelling Neglected Tropical Diseases diagnostics: the sensitivity of skin snips for Onchocerca volvulus in near elimination and surveillance settings. Parasit Vectors 9(1): 343. Bourguinat C., Pion S. D. S., Kamgno J., Gardon J., Duke B. O. L., Boussinesq M. and Prichard R. K. (2007). Genetic Selection of Low Fertile Onchocerca volvulus by Ivermectin Treatment. PLoS Negl Trop Dis 1(1). Bourguinat C., Ardelli B. F., Pion S. D., Kamgno J., Gardon J., Duke B. O., Boussinesq M. and Prichard R. K. (2008). P-glycoprotein-like protein, a possible genetic marker for ivermectin resistance selection in Onchocerca volvulus. Mol Biochem Parasitol 158(2): 101-111. Bradley J. E., Whitworth J. A. G. and Basáñez M. G. (2005). Onchocerciasis. Topley and Wilson’s Microbiology and Microbial Infections, 10th edition. Edward Arnold Publishers Ltd, London Parasitology: 781-801. Brattig N. W. (2004). Pathogenesis and host responses in human onchocerciasis: impact of Onchocerca filariae and Wolbachia endobacteria. Microbes Infect 6(1): 113-128. Burg R. W., Miller B. M., Baker E. E., Birnbaum J., Currie S. A., Hartman R., Kong Y. L., Monaghan R. L., Olson G., Putter I., Tunac J. B., Wallick H., Stapley E. O., Oiwa R. and Omura S. (1979). Avermectins, new family of potent anthelmintic agents: producing organism and fermentation. Antimicrob Agents Chemother 15(3): 361-367. Burnham G. (2007). Efficacy of ivermectin against Onchocerca volvulus in Ghana. Lancet 370(9593): 1125. Campbell W. C., Fisher M. H., Stapley E. O., Albers-Schonberg G. and Jacob T. A. (1983). Ivermectin: a potent new antiparasitic agent. Science 221(4613): 823-828. Campbell W. C. (2012). History of avermectin and ivermectin, with notes on the history of other macrocyclic lactone antiparasitic agents. Curr Pharm Biotechnol 13(6): 853- 865. CDC (2013a). Parasites - Onchocerciasis (also known as River Blindness). Available at: http://www.cdc.gov/parasites/onchocerciasis/biology.html, Updated on 21/05/2013; Accessed on 12/08/2015. CDC (2013b). Progress toward elimination of onchocerciasis in the Americas - 1993-2012. MMWR Morb Mortal Wkly Rep 62(20): 405-408. Chabala J. C., Mrozik H., Tolman R. L., Eskola P., Lusi A., Peterson L. H., Woods M. F., Fisher M. H., Campbell W. C., Egerton J. R. and Ostlind D. A. (1980).

140

University of Ghana http://ugspace.ug.edu.gh

Ivermectin, a new broad-spectrum antiparasitic agent. J Med Chem 23(10): 1134- 1136. Chavasse D. C., Post R. J., Lemoh P. A. and Whitworth J. A. (1992). The effect of repeated doses of ivermectin on adult female Onchocerca volvulus in Sierra Leone. Trop Med Parasitol 43(4): 256-262. Chavasse D. C., Post R. J., Davies J. B. and Whitworth J. A. (1993). Absence of sperm from the seminal receptacle of female Onchocerca volvulus following multiple doses of ivermectin. Trop Med Parasitol 44(3): 155-158. Cheeseman C. L., Delany N. S., Woods D. J. and Wolstenholme A. J. (2001). High- affinity ivermectin binding to recombinant subunits of the Haemonchus contortus glutamate-gated chloride channel. Mol Biochem Parasitol 114(2): 161-168. Cheke R. A. and Garms R. (2013). Indices of onchocerciasis transmission by different members of the Simulium damnosum complex conflict with the paradigm of forest and savanna parasite strains. Acta Trop 125(1): 43-52. Churcher T. S., Schwab A. E., Prichard R. K. and Basáñez M. G. (2008). An analysis of genetic diversity and inbreeding in Wuchereria bancrofti: implications for the spread and detection of drug resistance. PLoS Negl Trop Dis 2(4). Churcher T. S., Pion S. D., Osei-Atweneboana M. Y., Prichard R. K., Awadzi K., Boussinesq M., Collins R. C., Whitworth J. A. and Basáñez M. G. (2009). Identifying sub-optimal responses to ivermectin in the treatment of river blindness. Proc Natl Acad Sci U S A 106(39): 16716-16721. Coffeng L. E., Stolk W. A., Zouré H. G. M., Veerman J. L., Agblewonu K. B., Murdoch M. E., Noma M., Fobi G., Richardus J. H., Bundy D. A. P., Habbema D., de Vlas S. J. and Amazigo U. V. (2013). African Programme for Onchocerciasis Control 1995–2015: Model-estimated health impact and cost. PLoS Negl Trop Dis 7(1). Coffeng L. E., Stolk W. A., Zoure H. G., Veerman J. L., Agblewonu K. B., Murdoch M. E., Noma M., Fobi G., Richardus J. H., Bundy D. A., Habbema D., de Vlas S. J. and Amazigo U. V. (2014a). African Programme for Onchocerciasis Control 1995- 2015: updated health impact estimates based on new disability weights. PLoS Negl Trop Dis 8(6): e2759. Coffeng L. E., Stolk W. A., Hoerauf A., Habbema D., Bakker R., Hopkins A. D. and de Vlas S. J. (2014b). Elimination of African onchocerciasis: modeling the impact of increasing the frequency of ivermectin mass treatment. PLoS One 9(12): e115886. Conde C. and Cáceres A. (2009). Microtubule assembly, organization and dynamics in axons and dendrites. Available at: http://www.nature.com/nrn/journal/v10/n5/box/nrn2631_BX1.html; Accessed on 22/11/2016. Convit J., Schuler H., Borges R., Olivero V., Dominguez-Vazquez A., Frontado H. and Grillet M. E. (2013). Interruption of Onchocerca volvulus transmission in Northern Venezuela. Parasit Vectors 6(1): 289. Crosskey R. W. (1967). The classification of Simulium Latreille (Diptera: Simuliidae) from Australia, New Guinea and the Western Pacific. Journal of Natural History 1(1): Pages 23-51. Crosskey R. W. (1990). The Natural History of Blackflies. John Wiley & Sons Ltd, Chichester, England. Crump A. and Omura S. (2011). Ivermectin, 'wonder drug' from Japan: the human use perspective. Proc Jpn Acad Ser B Phys Biol Sci 87(2): 13-28.

141

University of Ghana http://ugspace.ug.edu.gh

Cruz-Ortiz N., Gonzalez R. J., Lindblade K. A., Richards F. O., Jr., Sauerbrey M., Zea- Flores G., Dominguez A., Oliva O., Catu E. and Rizzo N. (2012). Elimination of Onchocerca volvulus Transmission in the Huehuetenango Focus of Guatemala. J Parasitol Res 2012: 638429. Cupp E., Richards F., Lammie P. and Eberhard M. (2007). Efficacy of ivermectin against Onchocerca volvulus in Ghana. Lancet 370(9593): 1123; author reply 1124-1125. Cupp E. W., Duke B. O., Mackenzie C. D., Guzman J. R., Vieira J. C., Mendez-Galvan J., Castro J., Richards F., Sauerbrey M., Dominguez A., Eversole R. R. and Cupp M. S. (2004). The effects of long-term community level treatment with ivermectin (Mectizan) on adult Onchocerca volvulus in Latin America. Am J Trop Med Hyg 71(5): 602-607. Dadzie K. Y., Remme J., Baker R. H., Rolland A. and Thylefors B. (1990). Ocular onchocerciasis and intensity of infection in the community. III. West African rainforest foci of the vector Simulium sanctipauli. Trop Med Parasitol 41(4): 376- 382. Davidson A. L., Dassa E., Orelle C. and Chen J. (2008). Structure, Function, and Evolution of Bacterial ATP-Binding Cassette Systems. Microbiol Mol Biol Rev 72(2): 317-364. Davison A. C. and Hinkley D. V. (1996). Bootstrap methods and their application. Available at: http://statwww.epfl.ch/davison/BMA/CUPsample.pdf; Accessed on 01/02/2016. Diawara L., Traore M. O., Badji A., Bissan Y., Doumbia K., Goita S. F., Konate L., Mounkoro K., Sarr M. D., Seck A. F., Toe L., Touree S. and Remme J. H. F. (2009). Feasibility of Onchocerciasis Elimination with Ivermectin Treatment in Endemic Foci in Africa: First Evidence from Studies in Mali and Senegal. PLoS Negl Trop Dis 3(7). Diggle P. J., Heagerty P., Liang K. Y. and Zeger S. (2013). Analysis of longitudinal data. 2nd edition, Oxford, England: Oxford University Press. Dowell S. F., Sejvar J. J., Riek L., Vandemaele K. A., Lamunu M., Kuesel A. C., Schmutzhard E., Matuja W., Bunga S., Foltz J., Nutman T. B., Winkler A. S. and Mbonye A. K. (2013). Nodding syndrome. Emerg Infect Dis 19(9): 1374-1384. Dozie I. N., Onwuliri C. O., Nwoke B. E., Chukwuocha U. M. and Nwoke E. A. (2007). Prevalence of lymphatic complications due to onchocerciasis infection in Imo State, Nigeria. Nigerian Journal of Parasitology 27(1). Duerr H. P., Dietz K., Buttner D. W. and Schulz-Key H. (2001). A stochastic model for the aggregation of Onchocerca volvulus in nodules. Parasitology 123(Pt 2): 193-201. Duerr H. P., Dietz K., Schulz-Key H., Buttner D. W. and Eichner M. (2004). The relationships between the burden of adult parasites, host age and the microfilarial density in human onchocerciasis. Int J Parasitol 34(4): 463-473. Duke B. O. (1980). Observations on Onchocerca volvulus in experimentally infected chimpanzees. Tropenmed Parasitol 31(1): 41-54. Duke B. O., Zea-Flores G., Castro J., Cupp E. W. and Munoz B. (1990). Effects of multiple monthly doses of ivermectin on adult Onchocerca volvulus. Am J Trop Med Hyg 43(6): 657-664. Duke B. O., Zea-Flores G., Castro J., Cupp E. W. and Munoz B. (1992). Effects of three- month doses of ivermectin on adult Onchocerca volvulus. Am J Trop Med Hyg 46(2): 189-194.

142

University of Ghana http://ugspace.ug.edu.gh

Duke B. O., Marty A. M., Peett D. L., Gardo J., Pion S. D., Kamgno J. and Boussinesq M. (2002). Neoplastic change in Onchocerca volvulus and its relation to ivermectin treatment. Parasitology 125(Pt 5): 431-444. Eng J. K. and Prichard R. K. (2005). A comparison of genetic polymorphism in populations of Onchocerca volvulus from untreated- and ivermectin-treated patients. Mol Biochem Parasitol 142(2): 193-202. Eng J. K., Blackhall W. J., Osei-Atweneboana M. Y., Bourguinat C., Galazzo D., Beech R. N., Unnasch T. R., Awadzi K., Lubega G. W. and Prichard R. K. (2006). Ivermectin selection on beta-tubulin: evidence in Onchocerca volvulus and Haemonchus contortus. Mol Biochem Parasitol 150(2): 229-235. Engelbrecht F. and Schulz Key H. (1984). Observations on adult Onchocerca volvulus maintained in vitro. Transactions of the Royal Society of Tropical Medicine and Hygiene 78: 212-215. Evans D. S., Alphonsus K., Umaru J., Eigege A., Miri E., Mafuyai H., Gonzales-Peralta C., Adamani W., Pede E., Umbugadu C., Saka Y., Okoeguale B. and Richards F. O. (2014). Status of Onchocerciasis transmission after more than a decade of mass drug administration for onchocerciasis and lymphatic filariasis elimination in central Nigeria: challenges in coordinating the stop MDA decision. PLoS Negl Trop Dis 8(9): e3113. Fink D. L., Fahle G. A., Fischer S., Fedorko D. F. and Nutman T. B. (2011). Toward molecular parasitologic diagnosis: enhanced diagnostic sensitivity for filarial infections in mobile populations. J Clin Microbiol 49(1): 42-47. Foger K., Gora-Stahlberg G., Sejvar J., Ovuga E., Jilek-Aall L., Schmutzhard E., Kaiser C. and Winkler A. S. (2017). Nakalanga Syndrome: Clinical Characteristics, Potential Causes, and Its Relationship with Recently Described Nodding Syndrome. PLoS Negl Trop Dis 11(2): e0005201. Franz M. and Buttner D. W. (1983). The fine structure of adult Onchocerca volvulus. III. The cuticle, the interchordal hypodermis and the muscle cells of the female worm. Tropenmed Parasitol 34(1): 61-69. Frempong K. K., Walker M., Cheke R. A., Tetevi E. J., Gyan E. T., Owusu E. O., Wilson M. D., Boakye D. A., Taylor M. J., Biritwum N. K., Osei-Atweneboana M. and Basáñez M. G. (2016). Does Increasing Treatment Frequency Address Suboptimal Responses to Ivermectin for the Control and Elimination of River Blindness? Clin Infect Dis. Friedrich M. J. (2017). Potential Cause of Nodding Syndrome Identified. Jama 317(15): 1517. Gardon J., Boussinesq M., Kamgno J., Gardon-Wendel N., Demanga N. and Duke B. O. (2002). Effects of standard and high doses of ivermectin on adult worms of Onchocerca volvulus: a randomised controlled trial. Lancet 360(9328): 203-210. GHS M. (2007). Two-year strategic plan for integrated neglected tropical diseases control in Ghana. Available at: http://www.moh- ghana.org/UploadFiles/Publications/Plan%20for%20Pro- Poor%20Diseases120506091943.pdf; Accessed on: 15/10/15. Garms R., Lakwo T. L., Ndyomugyenyi R., Kipp W., Rubaale T., Tukesiga E., Katamanywa J., Post R. J. and Amazigo U. V. (2009). The elimination of the vector Simulium neavei from the Itwara onchocerciasis focus in Uganda by ground larviciding. Acta Trop 111(3): 203-210.

143

University of Ghana http://ugspace.ug.edu.gh

Haarbrink M., Terhell A. J., Abadi G. K., Mitsui Y. and Yazdanbakhsh M. (1999). Adverse reactions following diethylcarbamazine (DEC) intake in 'endemic normals', microfilaraemics and elephantiasis patients. Trans R Soc Trop Med Hyg 93(1): 91-96. Hall L. R. and Pearlman E. (1999). Pathogenesis of Onchocercal Keratitis (River Blindness). Clin Microbiol Rev 12(3): 445-453. Higazi T. B., Filiano A., Katholi C. R., Dadzie Y., Remme J. H. and Unnasch T. R. (2005). Wolbachia endosymbiont levels in severe and mild strains of Onchocerca volvulus. Mol Biochem Parasitol 141(1): 109-112. Higazi T. B., Zarroug I. M. A., Mohamed H. A., ElMubark W. A., Deran T. C. M., Aziz N., Katabarwa M., Hassan H. K., Unnasch T. R., Mackenzie C. D., Richards F. and Hashim K. (2013). Interruption of Onchocerca volvulus Transmission in the Abu Hamed Focus, Sudan. Am J Trop Med Hyg 89(1): 51-57. Hildebrandt J. C., Eisenbarth A., Renz A. and Streit A. (2012). Single worm genotyping demonstrates that Onchocerca ochengi females simultaneously produce progeny sired by different males. Parasitol Res 111(5): 2217-2221. Hildebrandt J. C., Eisenbarth A., Renz A. and Streit A. (2014). Reproductive biology of Onchocerca ochengi, a nodule forming filarial nematode in zebu cattle. Vet Parasitol 205(1-2): 318-329. Hoerauf A., Specht S., Buttner M., Pfarr K., Mand S., Fimmers R., Marfo-Debrekyei Y., Konadu P., Debrah A. Y., Bandi C., Brattig N., Albers A., Larbi J., Batsa L., Taylor M. J., Adjei O. and Buttner D. W. (2008). Wolbachia endobacteria depletion by doxycycline as antifilarial therapy has macrofilaricidal activity in onchocerciasis: a randomized placebo-controlled study. Med Microbiol Immunol 197(3): 295-311. Hotez P. J. (2007). Control of onchocerciasis-the next generation. Lancet 369: 1979-1980. Hougard J. M., Alley E. S., Yameogo L., Dadzie K. Y. and Boatin B. A. (2001). Eliminating onchocerciasis after 14 years of vector control: a proved strategy. J Infect Dis 184(4): 497-503. Hougard J. M., Alley E. S., Yaméogo L., Dadzie K. Y. and Boatin B. A. (2001). Eliminating onchocerciasis after 14 years of vector control: A proved strategy. Journal of Infectious Diseases 184: 497-503. Howard J. and Hyman A. A. (2003). Dynamics and mechanics of the microtubule plus end. Nature 422(6933): 753-758. Johnson T., Nutman T., Dowell S. and Nath A. (2017a). Closing the Loop between Nodding Syndrome and Onchocerca Infection. Trends Parasitol 33(7): 490. Johnson T. P., Tyagi R., Lee P. R., Lee M. H., Johnson K. R., Kowalak J., Elkahloun A., Medynets M., Hategan A., Kubofcik J., Sejvar J., Ratto J., Bunga S., Makumbi I., Aceng J. R., Nutman T. B., Dowell S. F. and Nath A. (2017b). Nodding syndrome may be an autoimmune reaction to the parasitic worm Onchocerca volvulus. Sci Transl Med 9(377). Junquera P. (2015). Parasites of Dogs, Cats, Horses & Livestock: Biology & Control. Available at: http://parasitipedia.net/index.php?option=com_content&view=article&id=2618&Ite mid=2897; Accessed on 27/08/2015. Kaiser C., Rubaale T., Tukesiga E., Kipp W. and Asaba G. (2015). Nodding Syndrome, Western Uganda, 1994. Am J Trop Med Hyg 93(1): 198-202. Kale O. O. (1978). A simplified technique for counting onchocercal microfilariae in skin snips. Bull World Health Organ 56(1): 133-137.

144

University of Ghana http://ugspace.ug.edu.gh

Kaplan R. M. (2004). Drug resistance in nematodes of veterinary importance: a status report. Trends Parasitol 20(10): 477-481. Katabarwa M., Lakwo T., Habomugisha P., Agunyo S., Byamukama E., Oguttu D., Ndyomugyenyi R., Tukesiga E., Ochieng G. O., Abwaimo F., Onapa A., Lwamafa D. W., Walsh F., Unnasch T. R. and Richards F. O. (2014). Transmission of Onchocerca volvulus by Simulium neavei in Mount Elgon focus of Eastern Uganda has been interrupted. Am J Trop Med Hyg 90(6): 1159-1166. Katabarwa M. N., Walsh F., Habomugisha P., Lakwo T. L., Agunyo S., Oguttu D. W., Unnasch T. R., Unoba D., Byamukama E., Tukesiga E., Ndyomugyenyi R. and Richards F. O. (2012). Transmission of onchocerciasis in wadelai focus of northwestern Uganda has been interrupted and the disease eliminated. J Parasitol Res 2012: 748540. Kerboeuf D. and Guégnard F. (2011). Anthelmintics Are Substrates and Activators of Nematode P Glycoprotein▿. Antimicrob Agents Chemother 55(5): 2224-2232. Klager S., Whitworth J. A., Post R. J., Chavasse D. C. and Downham M. D. (1993). How long do the effects of ivermectin on adult Onchocerca volvulus persist? Trop Med Parasitol 44(4): 305-310. Klager S. L., Whitworth J. A. and Downham M. D. (1996). Viability and fertility of adult Onchocerca volvulus after 6 years of treatment with ivermectin. Trop Med Int Health 1(5): 581-589. Klein C., Kuchler K. and Valachovic M. (2011). ABC proteins in yeast and fungal pathogens. Essays Biochem 50(1): 101-119. Kotze A. C., Hunt P. W., Skuce P., von Samson-Himmelstjerna G., Martin R. J., Sager H., Krucken J., Hodgkinson J., Lespine A., Jex A. R., Gilleard J. S., Beech R. N., Wolstenholme A. J., Demeler J., Robertson A. P., Charvet C. L., Neveu C., Kaminsky R., Rufener L., Alberich M., Menez C. and Prichard R. K. (2014). Recent advances in candidate-gene and whole-genome approaches to the discovery of anthelmintic resistance markers and the description of drug/receptor interactions. Int J Parasitol Drugs Drug Resist 4(3): 164-184. Kudzi W., Dodoo A. N. and Mills J. J. (2010). Genetic polymorphisms in MDR1, CYP3A4 and CYP3A5 genes in a Ghanaian population: a plausible explanation for altered metabolism of ivermectin in humans? BMC Med Genet 11: 111. Lamberton P. H., Cheke R. A., Walker M., Winskill P., Osei-Atweneboana M. Y., Tirados I., Tetteh-Kumah A., Boakye D. A., Wilson M. D., Post R. J. and Basáñez M. G. (2014). Onchocerciasis transmission in Ghana: biting and parous rates of host- seeking sibling species of the Simulium damnosum complex. Parasit Vectors 7: 511. Lamberton P. H., Cheke R. A., Winskill P., Tirados I., Walker M., Osei-Atweneboana M. Y., Biritwum N. K., Tetteh-Kumah A., Boakye D. A., Wilson M. D., Post R. J. and Basáñez M. G. (2015). Onchocerciasis transmission in Ghana: persistence under different control strategies and the role of the simuliid vectors. PLoS Negl Trop Dis 9(4): e0003688. Lamberton P. H., Cheke R. A., Walker M., Winskill P., Crainey J. L., Boakye D. A., Osei-Atweneboana M. Y., Tirados I., Wilson M. D., Tetteh-Kumah A., Otoo S., Post R. J. and Basanez M. G. (2016). Onchocerciasis transmission in Ghana: the human blood index of sibling species of the Simulium damnosum complex. Parasit Vectors 9(1): 432.

145

University of Ghana http://ugspace.ug.edu.gh

Leveque C. (1989). The Use of Insecticides in the Onchocerciasis Control Programme and Aquatic Monitoring in WestAfrica. Available at: http://globalecology.stanford.edu/SCOPE/SCOPE_38/SCOPE_38_5.5_Leveque_317- 336.pdf, Accessed on 05/01/2015. Li B. W., Rush A. C. and Weil G. J. (2014). High level expression of a glutamate-gated chloride channel gene in reproductive tissues of Brugia malayi may explain the sterilizing effect of ivermectin on filarial worms. Int J Parasitol Drugs Drug Resist 4(2): 71-76. Linton K. J. and Higgins C. F. (2007). Structure and function of ABC transporters: the ATP switch provides flexible control. Pflugers Arch 453(5): 555-567. Lipner E. M., Dembele N., Souleymane S., Alley W. S., Prevots D. R., Toe L., Boatin B., Weil G. J. and Nutman T. B. (2006). Field applicability of a rapid-format anti-Ov-16 antibody test for the assessment of onchocerciasis control measures in regions of endemicity. J Infect Dis 194(2): 216-221. Little M. P., Breitling L. P., Basáñez M. G., Alley E. S. and Boatin B. A. (2004a). Association between microfilarial load and excess mortality in onchocerciasis: an epidemiological study. Lancet 363(9420): 1514-1521. Little M. P., Basáñez M. G., Breitling L. P., Boatin B. A. and Alley E. S. (2004b). Incidence of blindness during the Onchocerciasis control programme in western Africa, 1971-2002. J Infect Dis 189(10): 1932-1941. Lobos E., Weiss N., Karam M., Taylor H. R., Ottesen E. A. and Nutman T. B. (1991). An immunogenic Onchocerca volvulus antigen: a specific and early marker of infection. Science 251(5001): 1603-1605. Lovato R., Guevara A., Guderian R., Proano R., Unnasch T., Criollo H., Hassan H. K. and Mackenzie C. D. (2014). Interruption of infection transmission in the onchocerciasis focus of Ecuador leading to the cessation of ivermectin distribution. PLoS Negl Trop Dis 8(5): e2821. Mackenzie C. D. (2007). Efficacy of ivermectin against Onchocerca volvulus in Ghana. Lancet 370(9593): 1123; author reply 1124-1125. Marty A. M. (2013). Dermatologic Manifestations of Filariasis. Available at: http://emedicine.medscape.com/article/1109642-overview#a2; Updated on 12/12/2013; Accessed on 12/08/2015. Moreno Y., Nabhan J. F., Solomon J., Mackenzie C. D. and Geary T. G. (2010). Ivermectin disrupts the function of the excretory-secretory apparatus in microfilariae of Brugia malayi. Proc Natl Acad Sci U S A 107(46): 20120-20125. Mossinger J. and Barthold E. (1987). Fecundity and localization of Dipetalonema viteae (Nematoda, Filarioidea) in the jird Meriones unguiculatus. Parasitol Res 74(1): 84- 87. Murdoch M. E., Hay R. J., Mackenzie C. D., Williams J. F., Ghalib H. W., Cousens S., Abiose A. and Jones B. R. (1993). A clinical classification and grading system of the cutaneous changes in onchocerciasis. Br J Dermatol 129(3): 260-269. Nana-Djeunga H., Bourguinat C., Pion S. D., Kamgno J., Gardon J., Njiokou F., Boussinesq M. and Prichard R. K. (2012). Single nucleotide polymorphisms in beta- tubulin selected in Onchocerca volvulus following repeated ivermectin treatment: possible indication of resistance selection. Mol Biochem Parasitol 185(1): 10-18. Nana-Djeunga H. C., Bourguinat C., Pion S. b. D., Bopda J., Kengne-Ouafo J. A., Njiokou F., Prichard R. K., Wanji S., Kamgno J. and Boussinesq M. (2014). Reproductive Status of Onchocerca volvulus after Ivermectin Treatment in an 146

University of Ghana http://ugspace.ug.edu.gh

Ivermectin-Naive and a Frequently Treated Population from Cameroon. PLoS Neglected Tropical Disease 8(4): e2824. Nelson G. S. (1958). Hanging groin and hernia complications of onchocerciasis. Trans R Soc Trop Med Hyg 52(3): 272-275. Noma M., Nwoke B. E., Nutall I., Tambala P. A., Enyong P., Namsenmo A., Remme J., Amazigo U. V., Kale O. O. and Seketeli A. (2002). Rapid epidemiological mapping of onchocerciasis (REMO): its application by the African Programme for Onchocerciasis Control (APOC). Ann Trop Med Parasitol 96 Suppl 1: S29-39. Noma M., Zouré H. G., Tekle A. H., Enyong P. A., Nwoke B. E. and Remme J. H. (2014). The geographic distribution of onchocerciasis in the 20 participating countries of the African Programme for Onchocerciasis Control: (1) priority areas for ivermectin treatment. Parasit Vectors 7: 325. O'Hanlon S. J., Slater H. C., Cheke R. A., Boatin B. A., Coffeng L. E., Pion S. D., Boussinesq M., Zoure H. G., Stolk W. A. and Basáñez M. G. (2016). Model-Based Geostatistical Mapping of the Prevalence of Onchocerca volvulus in West Africa. PLoS Negl Trop Dis 10(1): e0004328. Omade S. I., Omotunde A. T. and Gbenga A. S. (2015). Mathematical Modeling of River Blindness Disease with Demography Using Euler Method. Mathematical Theory and Modeling 5(5). Osei-Atweneboana M. Y., Eng J. K., Boakye D. A., Gyapong J. O. and Prichard R. K. (2007). Prevalence and intensity of Onchocerca volvulus infection and efficacy of ivermectin in endemic communities in Ghana: a two-phase epidemiological study. Lancet 369(9578): 2021-2029. Osei-Atweneboana M. Y., Awadzi K., Attah S. K., Boakye D. A., Gyapong J. O. and Prichard R. K. (2011). Phenotypic Evidence of Emerging Ivermectin Resistance in Onchocerca volvulus. PLoS Negl Trop Dis 5(3). Osei-Atweneboana M. Y., Boakye D. A., Awadzi K., Gyapong J. O. and Prichard R. K. (2012). Genotypic analysis of β-tubulin in Onchocerca volvulus from communities and individuals showing poor parasitological response to ivermectin treatment. Int J Parasitol Drugs Drug Resist 2: 20-28. Osei-Atweneboana M. Y., Lustigman S., Prichard R. K., Boatin B. A. and Basáñez M. G. (2012). A research agenda for helminth diseases of humans: health research and capacity building in disease-endemic countries for helminthiases control. PLoS Negl Trop Dis 6(4): e1602. Peelman F., Labeur C., Vanloo B., Roosbeek S., Devaud C., Duverger N., Denefle P., Rosier M., Vandekerckhove J. and Rosseneu M. (2003). Characterization of the ABCA transporter subfamily: identification of prokaryotic and eukaryotic members, phylogeny and topology. J Mol Biol 325(2): 259-274. Pilhofer M., Ladinsky M. S., McDowall A. W., Petroni G. and Jensen G. J. (2011). Microtubules in bacteria: Ancient tubulins build a five-protofilament homolog of the eukaryotic cytoskeleton. PLoS Biol 9(12): e1001213. Pion S. D., Nana-Djeunga H. C., Kamgno J., Tendongfor N., Wanji S., Njiokou F., Prichard R. K. and Boussinesq M. (2013). Dynamics of Onchocerca volvulus microfilarial densities after ivermectin treatment in an ivermectin-naive and a multiply treated population from Cameroon. PLoS Negl Trop Dis 7(2): e2084. Plaisier A. P., van Oortmarssen G. J., Remme J. and Habbema J. D. (1991). The reproductive lifespan of Onchocerca volvulus in West African savannah. Acta Trop 48(4): 271-284. 147

University of Ghana http://ugspace.ug.edu.gh

Pohl A., Devaux P. F. and Herrmann A. (2005). Function of prokaryotic and eukaryotic ABC proteins in lipid transport. Biochim Biophys Acta 1733(1): 29-52. Pooda H. S., Rayaisse J. B., Hien D. F., Lefevre T., Yerbanga S. R., Bengaly Z., Dabire R. K., Belem A. M., Sidibe I., Solano P. and Mouline K. (2015). Administration of ivermectin to peridomestic cattle: a promising approach to target the residual transmission of human malaria. Malar J 13 Suppl 1: 496. Post R. J. (1986). The cytotaxonomy of Simulium sanctipauli and Simulium soubrense (Diptera: Simuliidae) Genetica 69(3): 191-207. Post R. J., Cheke R. A., Boakye D. A., Wilson M. D., Osei-Atweneboana M. Y., Tetteh- Kumah A., Lamberton P. H., Crainey J. L., Yameogo L. and Basáñez M. G. (2013). Stability and change in the distribution of cytospecies of the Simulium damnosum complex (Diptera: Simuliidae) in southern Ghana from 1971 to 2011. Parasit Vectors 6: 205. Prasad R. and Goffeau A. (2012). Yeast ATP-binding cassette transporters conferring multidrug resistance. Annu Rev Microbiol 66: 39-63. Prichard R. K. (2007). Ivermectin resistance and overview of the Consortium for Anthelmintic Resistance SNPs. Expert Opin Drug Discov 2(s1): S41-52. Prost A. (1980). [Latency period in onchocerciasis]. Bull World Health Organ 58(6): 923- 925. R Core Team (2015). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available at: https://www.R- project.org/; Accessed on 23/08/2016. Rees D. C., Johnson E. and Lewinson O. (2009). ABC transporters: the power to change. Nat Rev Mol Cell Biol 10(3): 218-227. Remme J., Ba O., Dadzie K. Y. and Karam M. (1986). A force-of-infection model for onchocerciasis and its applications in the epidemiological evaluation of the Onchocerciasis Control Programme in the Volta River basin area. Bull World Health Organ 64(5): 667-681. Remme J. H., Amazigo U., Engels D., Barryson A. and Yameogo L. (2007). Efficacy of ivermectin against Onchocerca volvulus in Ghana. Lancet 370(9593): 1123-1124; author reply 1124-1125. Rodriguez-Perez M. A., Fernandez-Santos N. A., Orozco-Algarra M. E., Rodriguez- Atanacio J. A., Dominguez-Vazquez A., Rodriguez-Morales K. B., Real-Najarro O., Prado-Velasco F. G., Cupp E. W., Richards F. O., Jr., Hassan H. K., Gonzalez-Roldan J. F., Kuri-Morales P. A. and Unnasch T. R. (2015). Elimination of Onchocerciasis from Mexico. PLoS Negl Trop Dis 9(7): e0003922. Samba E. M. (1994). The Onchocerciasis Control Programme in West Africa. An example of effective public health management. Available at: http://whqlibdoc.who.int/pha/WHO_PHA_1.pdf>, Accessed on 18/10/2010. Sauerbrey M. (2008). The Onchocerciasis Elimination Program for the Americas (OEPA). Ann Trop Med Parasitol 102 Suppl 1: 25-29. Sauerbrey M., Rakers L. J. and Richards F. O. (2018). Progress toward elimination of onchocerciasis in the Americas. Int Health 10(suppl_1): i71-i78. Schulz-Key H., Albiez E. J. and Buttner D. W. (1977). Isolation of living adult Onchocerca volvulus from nodules. Tropenmed Parasitol 28(4): 428-430. Schulz-Key H., Jean B. and Albiez E. J. (1980). Investigations on female Onchocerca volvulus for the evaluation of drug trials. Tropenmed Parasitol 31(1): 34-40.

148

University of Ghana http://ugspace.ug.edu.gh

Schulz-Key H. and Karam M. (1986). Periodic reproduction of Onchocerca volvulus. Parasitol Today 2(10): 284-286. Schulz-Key H. (1988). The collagenase technique: how to isolate and examine adult Onchocerca volvulus for the evaluation of drug effects. Trop Med Parasitol 39 Suppl 4: 423-440. Schulz-Key H. (1990). Observations on the reproductive biology of Onchocerca volvulus. Acta Leiden 59(1-2): 27-44. Schulz-Key H. and Soboslay P. T. (1994). Reproductive biology and population dynamics of Onchocerca volvulus in the vertebrate host Parasite 1(1S): S53-S55. Shane L. (2005). Shane's simple guide to F-statistics. Available at: http://www.library.auckland.ac.nz/subject-guides/bio/pdfs/733Pop-g-stats2.pdf; Accessed on 20/12/2015. Specht S., Brattig N., Buttner M. and Buttner D. W. (2009). Criteria for the differentiation between young and old Onchocerca volvulus filariae. Parasitol Res 105(6): 1531- 1538. Stanton R. A., Gernert K. M., Nettles J. H. and Aneja R. (2011). Drugs That Target Dynamic Microtubules: A New Molecular Perspective. Med Res Rev 31(3): 443-481. Stingl P., Ross M., Gibson D. W., Ribas J. and Connor D. H. (1984). A diagnostic "patch test" for onchocerciasis using topical diethylcarbamazine. Trans R Soc Trop Med Hyg 78(2): 254-258. Stolk W. A., Walker M., Coffeng L. E., Basanez M. G. and de Vlas S. J. (2015). Required duration of mass ivermectin treatment for onchocerciasis elimination in Africa: a comparative modelling analysis. Parasit Vectors 8: 552. Tamarozzi F., Halliday A., Gentil K., Hoerauf A., Pearlman E. and Taylor M. J. (2011). Onchocerciasis: the role of Wolbachia bacterial endosymbionts in parasite biology, disease pathogenesis, and treatment. Clin Microbiol Rev 24(3): 459-468. Taylor H. R., Semba R. D., Newland H. S., Keyvan-Larijani E., White A., Dukuly Z. and Greene B. M. (1989a). Ivermectin treatment of patients with severe ocular onchocerciasis. Am J Trop Med Hyg 40(5): 494-500. Taylor H. R., Munoz B., Keyvan-Larijani E. and Greene B. M. (1989b). Reliability of detection of microfilariae in skin snips in the diagnosis of onchocerciasis. Am J Trop Med Hyg 41(4): 467-471. Taylor M. J., Awadzi K., Basáñez M. G., Biritwum N., Boakye D. A., Boatin B., Bockarie M., Churcher T. S., Debrah A., Edwards G., Hoerauf A., Mand S., Matthews G., Osei-Atweneboana M., Prichard R. K., Wanji S. and Adjei O. (2009). Onchocerciasis control: Vision for the future from a Ghanaian perspective. Parasites and Vectors 2(7). Tekle A. H., Elhassan E., Isiyaku S., Amazigo U. V., Bush S., Noma M., Cousens S., Abiose A. and Remme J. H. (2012). Impact of long-term treatment of onchocerciasis with ivermectin in Kaduna State, Nigeria: first evidence of the potential for elimination in the operational area of the African Programme for Onchocerciasis Control. Parasit Vectors 5: 28. ter Beek J., Guskov A. and Slotboom D. J. (2014). Structural diversity of ABC transporters. J Gen Physiol 143(4): 419-435. Thylefors B. and Rolland A. (1979). The risk of optic atrophy following suramin treatment of ocular onchocerciasis*. Bull World Health Organ 57(3): 479-480. Thylefors B. (2008). The Mectizan Donation Program (MDP). Annals of Tropical Medicine and Parasitology 102(Suppl 1): 39-44. 149

University of Ghana http://ugspace.ug.edu.gh

Tielsch J. M. and Beeche A. (2004). Impact of ivermectin on illness and disability associated with onchocerciasis. Trop Med Int Health 9(4): A45-56. Traore S., Wilson M. D., Sima A., Barro T., Diallo A., Ake A., Coulibaly S., Cheke R. A., Meyer R. R., Mas J., McCall P. J., Post R. J., Zoure H., Noma M., Yameogo L., Seketeli A. V. and Amazigo U. V. (2009). The elimination of the onchocerciasis vector from the island of Bioko as a result of larviciding by the WHO African Programme for Onchocerciasis Control. Acta Trop 111(3): 211-218. Traore M. O., Sarr M. D., Badji A., Bissan Y., Diawara L., Doumbia K., Goita S. F., Konate L., Mounkoro K., Seck A. F., Toe L., Toure S. and Remme J. H. (2012). Proof-of-principle of onchocerciasis elimination with ivermectin treatment in endemic foci in Africa: final results of a study in Mali and Senegal. PLoS Negl Trop Dis 6(9): e1825. Turner H. C., Osei-Atweneboana M. Y., Walker M., Tettevi E. J., Churcher T. S., Asiedu O., Biritwum N. K. and Basáñez M. G. (2013a). The cost of annual versus biannual community-directed treatment of onchocerciasis with ivermectin: Ghana as a case study. PLoS Negl Trop Dis 7(9): e2452. Turner H. C., Churcher T. S., Walker M., Osei-Atweneboana M. Y., Prichard R. K. and Basáñez M. G. (2013b). Uncertainty surrounding projections of the long-term impact of ivermectin treatment on human onchocerciasis. PLoS Negl Trop Dis 7(4): e2169. Turner H. C., Walker M., Churcher T. S. and Basáñez M. G. (2014a). Modelling the impact of ivermectin on River Blindness and its burden of morbidity and mortality in African Savannah: EpiOncho projections. Parasit Vectors 7: 241. Turner H. C., Walker M., Churcher T. S., Osei-Atweneboana M. Y., Biritwum N. K., Hopkins A., Prichard R. K. and Basáñez M. G. (2014b). Reaching the london declaration on neglected tropical diseases goals for onchocerciasis: an economic evaluation of increasing the frequency of ivermectin treatment in Africa. Clin Infect Dis 59(7): 923-932. Turner H. C., Walker M., Attah S. K., Opoku N. O., Awadzi K., Kuesel A. C. and Basáñez M. G. (2015). The potential impact of moxidectin on onchocerciasis elimination in Africa: an economic evaluation based on the Phase II clinical trial data. Parasit Vectors 8: 167. Turner J. D., Tendongfor N., Esum M., Johnston K. L., Langley R. S., Ford L., Faragher B., Specht S., Mand S., Hoerauf A., Enyong P., Wanji S. and Taylor M. J. (2010). Macrofilaricidal activity after doxycycline only treatment of Onchocerca volvulus in an area of Loa loa co-endemicity: a randomized controlled trial. PLoS Negl Trop Dis 4(4): e660. Ubachukwu P. O. (2006). Socio-economic impact of onchocerciasis with particular reference to females and children: A review. Animal Research International 3(2): 494-504. Ughachukwu P. and Unekwe P. (2012). Efflux Pump-Mediated Resistance in Chemotherapy. Ann Med Health Sci Res 2(2): 191-198. Vajime C. G. and Dunbar R. W. (1975). Chromosomal identification of eight species of the subgenus Edwardsellum near and including Simulium (Edwardsellum) damnosum Theobald (Deptera: Simuliidae). Tropenmed Parasitol 26(1): 111-138. Vajime C. G. (1989). Cytotaxonomy of Sirba form populations of the Simulium damnosum complex in west Africa: amendments to sex chromosomes and sibling status. Trop Med Parasitol 40(4): 464-467.

150

University of Ghana http://ugspace.ug.edu.gh

Vercruysse J. (2014). Macrocyclic Lactones; The Merck Verterinary Manual. Available at: http://www.merckvetmanual.com/mvm/pharmacology/anthelmintics/macrocyclic_lacto nes.html; Reviewed in September 2014; Accessed on 23/08/2015. Wagbatsoma V. A. and Okojie O. H. (2004). Psychosocial effects of river blindness in a rural community in Nigeria. J R Soc Promot Health 124(3): 134-136. Walker M., Little M. P., Wagner K. S., Soumbey-Alley E. W., Boatin B. A. and Basáñez M. G. (2012). Density-Dependent Mortality of the Human Host in Onchocerciasis: Relationships between Microfilarial Load and Excess Mortality. PLoS Negl Trop Dis 6(3). Walker M., Churcher T. S. and Basáñez M. G. (2014). Models for measuring anthelmintic drug efficacy for parasitologists. Trends Parasitol 30(11): 528-537. Walker M., Specht S., Churcher T. S., Hoerauf A., Taylor M. J. and Basáñez M. G. (2015). Therapeutic efficacy and macrofilaricidal activity of doxycycline for the treatment of river blindness. Clin Infect Dis 60(8): 1199-1207. Walker M., Pion S. D. S., Fang H., Gardon J., Kamgno J., Basáñez M. G. and Boussinesq M. (2017). Macrofilaricidal Efficacy of Repeated Doses of Ivermectin for the Treatment of River Blindness. Clin Infect Dis 65(12): 2026-2034. Wanji S., Tendongfor N., Nji T., Esum M., Che J. N., Nkwescheu A., Alassa F., Kamnang G., Enyong P. A., Taylor M. J., Hoerauf A. and Taylor D. W. (2009). Community-directed delivery of doxycycline for the treatment of onchocerciasis in areas of co-endemicity with loiasis in Cameroon. Parasit Vectors 2(1): 39. Weil G. J., Steel C., Liftis F., Li B. W., Mearns G., Lobos E. and Nutman T. B. (2000). A rapid-format antibody card test for diagnosis of onchocerciasis. J Infect Dis 182(6): 1796-1799. Wessler J. D., Grip L. T., Mendell J. and Giugliano R. P. (2013). The P-glycoprotein transport system and cardiovascular drugs. J Am Coll Cardiol 61(25): 2495-2502. West S., Munoz B. and Sommer A. (2013). River blindness eliminated in Colombia. Ophthalmic Epidemiol 20(5): 258-259. WHO (1995). Onchocerciasis and its control. Report of a WHO Expert Committee on Onchocerciasis Control. World Health Organ Tech Rep Ser 852: 1-104. WHO (1998). Community-Directed Treatment with Ivermectin (CDTI). Available at: http://www.who.int/apoc/publications/cdti_practical_guide_for_trainers_of_cdds.pdf? ua=1; Accessed on 21 June, 2016. WHO (2008). Status of onchocerciasis in APOC countries. Available at: http://www.who.int/apoc/onchocerciasis/status/en/index.html. Accessed on 05/10/2010. WHO (2009). A major cause of blindness and poverty - especially in Africa Onchocerciasis/River Blindness. Available at: http://www.who.int/blindness/Vision2020_repor t.pdf . Accessed on 05/10/2010. WHO (2012a). Accelerating work to overcome the global impact of neglected tropical diseases; A roadmap for implementation. Available at: http://www.who.int/neglected_diseases/NTD_RoadMap_2012_Fullversion.pdf. Accessed on 01/02/2016. WHO (2012b). Final Communique; 18th Session of the Joint Action Forum, Bujumbura, Burundi. Available at: http://www.who.int/apoc/about/structure/jaf/Final_Communique_JAF_18_English_fin al_with_annexes.pdf; Accessed on 22/11/2016. 151

University of Ghana http://ugspace.ug.edu.gh

WHO (2015a). Onchocerciasis. Fact sheet No. 374, http://www.who.int/mediacentre/factsheets/fs374/en/, updated on March, 2015; Accessed on 12/08/2015. WHO (2015b). Prevention of Blindness and Visual Impairment; Causes of blindness and visual impairment. www.who.int/blindness/causes/en/ ; Accessed on 17/08/2015. Wilson M. D., Cheke R. A., Flasse S. P., Grist S., Osei-Ateweneboana M. Y., Tetteh- Kumah A., Fiasorgbor G. K., Jolliffe F. R., Boakye D. A., Hougard J. M., Yameogo L. and Post R. J. (2002). Deforestation and the spatio-temporal distribution of savannah and forest members of the Simulium damnosum complex in southern Ghana and south-western Togo. Trans R Soc Trop Med Hyg 96(6): 632-639. Wogu M. D. and Okaka C. E. (2008). Prevalence and socio-economic effects of onchocerciasis in Okpuje, Owan West Local Government Area, Edo State, Nigeria. . International Journal of Biomedical & Health Sciences 4(3). Wolstenholme A. J. and Rogers A. T. (2005). Glutamate-gated chloride channels and the mode of action of the avermectin/milbemycin anthelmintics. Parasitology 131(Suppl): S85–S95.

152

University of Ghana http://ugspace.ug.edu.gh

APPENDICES

Appendix I: Consent form and ethical clearance certificates

COUNCIL FOR SCIENTIFIC AND INDUSTRIAL RESEARCH INSTITUTIONAL REVIEW BOARD

INFORMED CONSENT FORM

PARTICIPANT INFORMATION AND CONSENT FORM FOR ONCHOCERCIASIS EPIDEMIOLOGICAL STUDY

PROJECT TITLE: DEVELOPMENT OF DIAGNOSTIC GENETIC MARKERS TO DETECT SUB-OPTIMAL RESPONSE TO IVERMECTIN

Investigators:

Dr Mike Yaw Osei-Atweneboana: Principal Investigator Council for Scientific and Industrial Research Accra, Ghana.

Prof. Warwick Grant: Co- Investigator Latrobe University, Melbourne, Australia:

Adjami Aime Gilles: Co- Investigator APOC, Burkina-Faso

153

University of Ghana http://ugspace.ug.edu.gh

General Information about Research Human onchocerciasis is a disease that causes disability usually from troublesome itching, skin disease and loss of vision, which may lead to total blindness. The disease is caused by a worm that lives inside the body of infected people and is carried from one person to another by small biting flies known as blackflies. The worms grow in the body, become adults and are housed in nodules located inside your body or under the skin. The nodules contain the adult worms which produce millions of very small baby worms (called microfilariae) which invade the skin and the eye. The microfilariae in the skin cause itching and colour changes of the skin. Onchocerciasis is still a problem in Ghana, and is controlled by giving a drug known as ivermectin to every person living in areas where the disease presents a problem. Ivermectin kills microfilariae, but does not kill the adult worms. In Ghana, many endemic communities have received between 10 and 25 annual rounds of ivermectin, yet the disease is still present in some communities. This means that in addition to the distribution of ivermectin we need to find other ways to help control the disease.

The purpose of the study is to find ways to monitor if ivermectin is still working very well and killing the worms in the body as it used to do some 10 to 20 years ago. Where ivermectin may not be working very well, we also want to see whether we can find something in the worm or in the body of the people that do not allow ivermectin to work well like some 20 years ago. To do this, we will need people like you, to volunteer to have the worms in your body measured at different times of the year and also give us information on how often you have been treated. We will take two small pieces of skin (a size smaller than a match head) around your waist. The skin sample will be observed under the microscope to measure the worms in your body and how these numbers change between treatments. We will remove nodules from your body to examine the worms, after that a nurse will come to your home every other day to see you and ensure everything is going well with the healing of the wound.

The worm samples will be used to find out what is in the worm that makes ivermectin works well or does not work well. This will help us find a way of monitoring the usefulness of the drug for onchocerciasis control. This research will also inform the Ghana Health Services as to whether providing ivermectin alone has the best effect on worms in all communities or we 154

University of Ghana http://ugspace.ug.edu.gh

will need additional drug for some areas where ivermectin is not working well like some 20 years ago.

Possible Risks and Discomforts Taking two small pieces of skin (a size smaller than a match head) will feel like a pinch. Normally there will not be any blood, but you may experience a little discomfort or slight pain during the procedure. The sites from where we take the snips should be kept clean and dry for as long as possible. When the medicine that will be injected before the removal of the nodule stops working (around 1 to 2 hours after the nodules have been taken out), you may feel some pain. You will be provided with some pain killers to reduce the pain, anti-inflammatory drugs and antibiotics to prevent any infection after nodule removal. A nurse will come to your home every other day to see you and ensure everything is going well with the healing of the wound. In case of complications with wound healing, arrangements will be made for you to report to a nearby hospital.

Possible Benefits You will not have any direct benefits from participating in this research; however, the removal of the nodules will reduce the number of worm in your body. The study will not directly help you to get better, but you will still continue to take ivermectin which will help you if you have the worms in your body. Your participation will benefit you and the future generations of children and adults in Ghana by helping us to produce methods of monitoring how best the drug is working so that better intervention will be used to eliminate the disease.

Alternatives to Participation If you decide not to participate in this study, you will still receive the regular annual or semi- annual ivermectin treatment which is being offered by the Ghana Health Services. Also medical care available to you in your respective community or village will not be affected or compromised.

155

University of Ghana http://ugspace.ug.edu.gh

Confidentiality Once you agree to participate, we will assign you a code number. We will be very careful about the information that we have collected from you and we will make absolutely sure that when we tell people about our findings, no-one will be able to know your personal identity.

Compensation You will not be compensated for the samples you provide, however, for the time loss by not going to farm/work during the sample collection day, you will receive a day’s work replacement allowance of five Ghana cedis (Gh¢ 5.00). Where nodules will be removed, you will receive four days’ work replacement allowance of twenty Ghana cedis (Gh¢ 20.00)

Voluntary Participation and Right to Leave the Research Before giving your consent, by signing this document, the methods, inconveniences, risks and benefits, and alternatives will be explained and your questions answered to your satisfaction. You may end your participation in this study at any time. Also your participation in this study may be ended by the investigators for reasons that will be explained to you. New information that develops during the study will be given to you, especially if it may affect your willingness to continue with the study. You do not give up any legal rights by signing this document. You can obtain further information from Council for Scientific and Industrial Research (CSIR) – Water Research Institute (+233-302-775351) or Dr. Mike Osei-Atweneboana (at 0203176771). The Institutional Review Board of the Council for Scientific and Industrial Research, Accra, Ghana, have reviewed this study, evaluated the potential risks and benefits and have granted approval to allow us to solicit participants in this study.

Notification of Significant New Findings After this study, we will analyse the data in Ghana. We will communicate these results as we go along in the project. We will let you know of our results when we revisit the communities later or through the Council for Scientific and Industrial Research, or through the community ivermectin distributors.

156

University of Ghana http://ugspace.ug.edu.gh

Contacts for Additional Information If you have any further questions please contact Dr Mike Yaw Osei-Atweneboana of the Council for Scientific and Industrial Research-Water Research Institute, Accra, and he will be able to explain further to you what our study is about. You should also contact this person if you would like to withdraw from the study, or if you have any worries regarding your participation in this study. The contact address is as follows:

Dr Mike Y. Osei-Atweneboana Department of Environmental Biology and Health Council for Scientific and Industrial Research- Water Research Institute. P.O. Box M 32, Accra Ghana, Mobile: 0203176771

Your rights as a Participant This research has been reviewed and approved by the Institutional Review Board of the Council for Scientific and Industrial Research (CSIR-IRB). If you have any questions about your rights as a research participant you can contact the IRB Office between the hours of 8am-5pm through the landline 0302777651 (extension number 1002) or email addresses: [email protected]. You may also contact the Chairman of the CSIR-IRB, through mobile number 0204362635 when necessary. A copy of the written information sheet and of the signed Informed Consent form will be given to you to keep.

157

University of Ghana http://ugspace.ug.edu.gh

Volunteer Agreement I confirm that I have read / been read the Participant Information Sheet explaining the research Project on “Development of Diagnostic Genetic Markers to Detect Sub-Optimal Response to Ivermectin”. I understand that the researchers are asking me to participate in their study by having skin snips samples taken from me as well as removal of nodules at in the study. I have been given a chance to ask questions until I feel that all of my questions have been answered. I know that my participation is entirely voluntary and that this will have no consequences for my usual participation in community-based ivermectin treatment. I also know that I can withdraw my consent at any time in the future.

______Name Date Signature or mark of volunteer

______Name Date Signature of Witness

______Name Date Signature of Person taking Consent

158

University of Ghana http://ugspace.ug.edu.gh

159

University of Ghana http://ugspace.ug.edu.gh

160

University of Ghana http://ugspace.ug.edu.gh

Appendix II: Estimation of community microfilarial loads, prevalence and confidence intervals; Marginal regression models

A. Community Microfilarial Load and Community Microfilarial Prevalence

The community microfilarial loads (CMFL) for each community was estimated by taking the arithmetic mean of the two microfilarial counts (from the right and left iliac crest snips) recorded per individual i , denoted mi , adding 1 to each of these values, and then taking the geometric mean across all individuals aged 20 years, then subtracting 1 from the geometric mean. The formula used is shown below:

 ni  ln mi  )1(  CMFL exp  i1  1  n   

NB: n is the number of individuals aged  20 years that were skin-snipped within a community at a specific sampling time.

The community microfilarial prevalence (CMFP) for each community was estimated by converting the two microfilarial counts per individual into a binary variable of infected or not infected (where infected represents being positive for microfilariae in either or both snips), denoted pi , and then taking the arithmetic mean of these values as shown below:

ni  pi CMFP i1 . n

B. Confidence interval estimation by bootstrapping The 95% confidence intervals (CIs) associated with the CMFL and CMFP were estimated using a non-parametric bootstrap resampling technique (Davison and Hinkley, 1996). The bootstrap resampling was performed as follows:

161

University of Ghana http://ugspace.ug.edu.gh

i) A number of participants (n ) with individual mean microfilarial counts (mi ) or

individual binary indicator of infection (pi ) were re-sampled (at random) with replacement. ii) These individual re-sampled (new sample) or were used to recalculate new CMFL or CMFP using the CMFL and CMFP formulae shown above. iii) Steps (i) and (ii) were repeated 10,000 times to yield an empirical sampling distribution of CMFL and CMFP. iv) The confidence intervals (CIs) were calculated from the 2.5th and 97.5th percentiles of the CMFL and CMFP sampling distributions.

C. Marginal regression models Marginal models are regression models that are suitable for analysing correlated data, i.e. in this case, correlation among repeated microfilarial counts measured from the same person. Two types of regression models were defined in this study (Model 1 and Model 2 in Table 4.1, Chapter 4) which were fitted to data on the microfilarial counts measured from the 217 longitudinal cohort followed up from baseline (July 2010) to six months after the second biannual treatment in July 2011 (Figure 3.2). The models were used to make predictions on the microfilarial counts at time points where participants were not sampled (e.g. in October 2010; see Figure 4.3) NB: Microfilarial counts = dependent variable Sampling time = covariate Community = covariate Age = covariate Sex = covariate

The microfilarial counts from model output were used to estimate the repopulation rates in this study (the microfilarial repopulation rate here is defined as the mean number of microfilariae counts expressed as a percentage of the previous mean–before treatment with ivermectin).

Model 1: The sampling time was defined as a categorical variable (i.e. at day 0, 90 or 180 after ivermectin treatment). In order to prevent the same value of repopulation rates generated 162

University of Ghana http://ugspace.ug.edu.gh

from the model (software), there was an interaction term allowed between sampling time and community. This enabled repopulation rates (microfilarial counts) to vary among communities. The model was constructed such that the mean number of microfilariae at a particular sampling time could be expressed as a percentage of the mean number immediately before the preceding round of ivermectin treatment (i.e. either in July 2010 or January 2011, Figure 3.2) by taking the exponent (because the models used were log-linear) of the sum of the relevant covariate coefficients.

Two variants of Model 1 were defined based on sampling time: i) Model 1A: This model ignored the exact number of days since the preceding round of ivermectin treatment. ii) Model 1B: This model incorporated the exact number of days since the preceding round of ivermectin treatment as an offset associated with the relevant sampling time (Table 4.1). By adjusting for the exact number of days since the preceding treatment, it enabled an estimation of six-monthly repopulation rates that were directly comparable in each community; one for each of the two consecutive periods of microfilarial repopulation (July 2010 to January 2011 and January 2011 to July 2011).

Model 2: The time since the preceding ivermectin treatment was defined as a continuous variable interacting with community. Like Model 1, this permitted rates of repopulation to vary among communities, but unlike Model 1, a single community-specific rate of microfilarial repopulation was estimated, informed by the data collected during both repopulation periods. Here, treatment round was also incorporated as a categorical variable interacting with time since the preceding treatment. This captured non community-specific heterogeneity/variation in microfilarial repopulation rates between the consecutive repopulation periods (i.e. assumption that variation in repopulation rates from July 2010 to January 2011 and January 2011 to July 2011 in all the communities were the same) that affects all communities are approximately equally. Such variation may arise, for example, from cumulative effects on the fertility of O. volvulus after repeated frequent exposures to ivermectin (Gardon et al., 2002).

163

University of Ghana http://ugspace.ug.edu.gh

Mathematically, both Model 1 and Model 2 have a similar structure, with a systematic component specified as:

mg )(  ijij βx j j = where mij is the expected value of microfilarial count (for 1,2 ) from individual i (for i = 1,2,...,n); xij is a vector (collection) of covariates i.e. time, community, age and sex, and

β is an accompanying vector of regression coefficients. The covariates within include a categorical variable indicating the time when a microfilarial count was measured (i.e. at day 0, 90 or 180 after ivermectin treatment); the community from which a count was measured; the additive covariates of age group [0, 20], (20, 40], (40, 60] and (60, 82] and sex, and the interactions between time and community. The inclusion of additive stratum adjustments for age and sex means that the mean (expected) number of microfilariae can vary among these strata.

The interactions between sampling date and community ensured that rates of repopulation could vary among communities, but constant within community strata. The difference between Model 1 and Model 2 arises in the specific construction of the interaction term, namely, in Model 1, time as a categorical variable is excluded (Model 1A) or included (Model 1B) an offset for the exact number of days since the preceding round of ivermectin treatment. But in Model 2, time is a continuous variable indicating the exact number of days since the preceding round of ivermectin treatment. Since the longitudinal cohort was followed up at only 3 and 6 months after the preceding ivermectin treatment, the underlying assumption of Model 2 (with time as a continuous covariate) is that numbers of microfilariae between these times (3 and 6 months) increase approximately log-linearly. This assumption appeared to capture adequately in the trend as shown in Figure 4.3.

D. Estimation of variance

The variance of both model types, mv ij )( , was specified as a linear function of the mean, permitting extra-Poisson variation (overdispersion):

ij )(   mmv ij

164

University of Ghana http://ugspace.ug.edu.gh

Where  is an estimated scale parameter. The correlation structure of the repeated measures was assumed to be exchangeable among microfilarial count measures from the same participant at the same time. That is, a single correlation parameter, 1 , was used to define the ‘cross-sectional’ correlation among repeated measures made at day 0—before treatment— or at days 90 or 180—after treatment. The correlation among repeated measures made at different times, for example among microfilarial counts from the same individual at day 0 and day 90 was given a separate, ‘longitudinal’ correlation parameter 2 . The models were fitted to the data using generalized estimating equations implemented with the geepack package for R (R Core Team, 2015).

165

University of Ghana http://ugspace.ug.edu.gh

APPENDIX III: Community Microfilarial Load of Onchocerca volvulus at Different Time Points

Community Microfilarial Load (CMFL) of Onchocerca volvulus in 2004 (Osei-Atweneboana et al., 2007), 2010 (this study, before introduction of biannual ivermectin treatment) and 2013 (this study, after 4 or 5 rounds of biannual ivermectin treatment) in 10 sentinel communities of the Neglected Tropical Disease Programme of the Ghana Health Service. The CMFL is defined as the geometric mean number of microfilariae (including zero counts) per skin snip in people aged ≥20 years.

Community Community Microfilarial Load Oct 2004a Jul 2010 % change Mar 2013 Jun 2013 % change Jul (95% CI) Oct 2004 to (95% CI) (95% CI) 2010 to Jul 2010 Mar/Jun 2013 Agborlekame 1 NA 1.04 NA 0.31 NA − 70% (0.56, 1.73) (0.11, 0.62) Asubende 0.62 0.33 − 47% NA 0.17 − 48% (0.13, 0.58) (0.03, 0.38) Baaya 0.28 0.00 NC NA 0.03 NC (0.00, 0.01) (0.00, 0.07) Jagbenbendo 2.12 1.09 − 49% 0.58 NA − 47% (0.74, 1.56) (0.33, 0.94) Kyingakrom 2.85 0.32 − 89% NA 0.35 + 10% (0.15, 0.57) (0.09, 0.77) New-Longoro 1.42 0.20 − 86% 0.01 NA − 95% (0.093, 0.33) (0.00, 0.04) Ohiampe 0.21 0.091 − 57% NA 0.15 + 65% (0.02, 0.19) (0.01, 0.36) Senyase 0.36 0.11 − 69% NA 0.04 − 64% (0.03, 0.21) (0.00, 0.12) Takumdo NA 1.57 NA 0.17 NA − 89% (1.01, 2.29) (0.03, 0.37) Wiae 1.20 0.35 − 71% 0.10 NA − 71% (0.19, 0.56) (0.03, 0.19) a Estimates presented by Osei-Atweneboana et al., (2007); CI = confidence interval, estimated using the numerical bootstrap approach (Appendix II B), NA = not available, community not studied at that time, NC = not calculated, only one individual positive for microfilariae in 2010 and 2013

166

University of Ghana http://ugspace.ug.edu.gh

APPENDIX IV: Community Microfilarial Prevalence of Onchocerca volvulus at Different Time Points

Community Microfilarial Prevalence (CMFP) of Onchocerca volvulus in 2004 (Osei- Atweneboana et al., 2007), 2010 (this study, before introduction of biannual ivermectin treatment) and 2013 (this study, after 4 or 5 rounds of biannual ivermectin treatment) in 10 sentinel communities of the Neglected Tropical Disease Programme of the Ghana Health Service. The CMFP is define as the prevalence of microfilariae in people aged ≥20 years in the community

Community Community Microfilarial Prevalence Oct Jul 2010 % change Mar 2013 Jun 2013 % change 2004a (95% CI) Oct 2004 to (95% CI) (95% CI) Jul 2010 to Jul 2010 Mar/Jun 2013 Agborlekame 1 NA 34.6% NA 17.6% NA − 49% (24.6%, 47%) (6.1%, 29.3%) Asubende 13.9% 20.1% + 45% NA 15.5% − 23% (8.5%, 33%) (2.9%, 29.7%) Baaya 8.7% 0.5% NC NA 2.4% NC (0.0%, 1.6%) (0.0%, 5.8%) Jagbenbendo 43.3% 38.3% − 12% 25.7% NA − 33% (29.6%, (16.8%, 36.3%) 47.5%) Kyingakrom 50.8% 12.6% − 75% NA 12.5% − 0.8% (7.0%, 19.4%) (4.5%, 22.9%) New-Longoro 35.8% 9.8% − 73% 2.5%e NA − 74% (5.7%, 14.7%) (0.0%, 6.3%) Ohiampe 5.0% 4.2% − 16% NA 4.9% + 17% (0.82%, 8.5%) (1.2%, 10.0%) Senyase 4.3% 9.1% + 112% NA 3.4% − 63% (3.3%, 16%) (0.0%, 8.8%) Takumdo NA 37.8% NA 6.6% NA − 83% (29.3%, (2.2%, 12.4%) 47.0%) Wiae 38.5% 12.1% − 69% 5.9% NA − 51% (7.7%, 16.7%) (1.7%, 10.0%) a Estimates presented by Osei-Atweneboana et al., (2007); CI = confidence interval, estimated using the numerical bootstrap approach (Appendix II B), NA = not available, community not studied at that time, NC = not calculated, only one individual positive for microfilariae in 2010 and 2013

167

University of Ghana http://ugspace.ug.edu.gh

APPENDIX V: Six-month Onchocerca volvulus Microfilarial Skin Repopulation Rates at Different Time Points

Six-month Onchocerca volvulus microfilarial skin repopulation rates in 2004 (Osei- Atweneboana et al., 2007) and 2010–2011 (this study, after introduction of biannual ivermectin treatment) in 10 sentinel communities of the Neglected Tropical Disease Programme of the Ghana Health Service. Six-month repopulation rate is define as the mean number of microfilariae per participant six months after a round of ivermectin treatment expressed as a percentage of the mean number of microfilariae just before the preceding round of treatment.

Community Six-month microfilarial skin repopulation ratesa Nov 2004 – Apr 2005b Jul 2010 – Jan 2011 Jan 2011 – Jul 2011 (95% CI) (95% CI) Agborlekame 1 NAd 50.1% 26.1% (26.5%, 94.9%) (10.6%, 64%) Asubende 22.1% 53.8% 129% (26%, 112%) (48.9%, 250%)f Baaya 12.3% NRe NR Jagbenbendo 36.3% 46.2% 17.9% (28%, 76.2%) (7.2%, 44.8%) Kyingakrom 53.8% 69.6% 94.3% (34%, 143%) (40.8%, 218%) New-Longoro 22.5% 41.6% 74.3% (16.4%, 105%) (26.1%, 211%) Ohiampe 16.2% 54.3% 42.2% (31%, 94.9%) (7.7%, 231%) Senyase 16.0% 45.8% 53.4% (17.6%, 120%) (12.6%, 227%) Takumdo NAd 49.4% 22.1% (30.7%, 79.4%) (10.5%, 46.6%) Wiae 29.6% 44.4% 25.2% (22%, 89.8%) (5.6%, 114%) a b estimates from data presented by Osei-Atweneboana et al. (2007); CI = confidence interval, estimated using the numerical bootstrap approach (Appendix II B); NA = not available, community not studied at that time; NR = not reported due to insufficient sample size and excessively large associated uncertainties; f upper bound truncated at 250%.

168

University of Ghana http://ugspace.ug.edu.gh

Appendix VI: Chromatograms of DNA sequence showing SNP positions

Figure A: Chromatogram showing SNP within beta-tubulin gene at position 1183. The first sequence at the upper part of the figure is the reference sequence and the rest of the sequences below it represent the samples. The chromatograms of the respective sequences are shown at the lower part of the figure. Each row of the chromatogram represent a sample. The different coloured peaks are the various nucleotides as follows: Green = Adenine (A), Blue = Cytosine (C), Black = Guanine (G) and red = Thymine (T). The black boxes vertically aligned indicate the SNP position (1183). Two peaks at the same position indicate heterozygosity at that locus of the gene. Where there is a complete change of nucleotide, there is a replacement of different nucleotide compared to the reference sequence.

169

University of Ghana http://ugspace.ug.edu.gh

Figure B: Chromatogram showing SNP within beta-tubulin gene at position 1188. The first sequence at the upper part of the figure is the reference sequence and the rest of the sequences below it represent the samples. The chromatograms of the respective sequences are shown at the lower part of the figure. Each row of the chromatogram represent a sample. The different coloured peaks are the various nucleotides as follows: Green = Adenine (A), Blue = Cytosine (C), Black = Guanine (G) and red = Thymine (T). The black boxes vertically aligned indicate the SNP position (1188). Two peaks at the same position indicate heterozygosity at that locus of the gene. Where there is a complete change of nucleotide, there is a replacement of different nucleotide compared to the reference sequence.

170

University of Ghana http://ugspace.ug.edu.gh

Figure C: Chromatogram showing SNP within beta-tubulin gene at position 1298. The first sequence at the upper part of the figure is the reference sequence and the rest of the sequences below it represent the samples. The chromatograms of the respective sequences are shown at the lower part of the figure. Each row of the chromatogram represent a sample. The different coloured peaks are the various nucleotides as follows: Green = Adenine (A), Blue = Cytosine (C), Black = Guanine (G) and red = Thymine (T). The black boxes vertically aligned indicate the SNP position (1298). Two peaks at the same position indicate heterozygosity at that locus of the gene. Where there is a complete change of nucleotide, there is a replacement of different nucleotide compared to the reference sequence.

171

University of Ghana http://ugspace.ug.edu.gh

Figure D: Chromatogram showing SNP within beta-tubulin gene at position 1308. The first sequence at the upper part of the figure is the reference sequence and the rest of the sequences below it represent the samples. The chromatograms of the respective sequences are shown at the lower part of the figure. Each row of the chromatogram represent a sample. The different coloured peaks are the various nucleotides as follows: Green = Adenine (A), Blue = Cytosine (C), Black = Guanine (G) and red = Thymine (T). The black boxes vertically aligned indicate the SNP position (1308). Two peaks at the same position indicate heterozygosity at that locus of the gene. Where there is a complete change of nucleotide, there is a replacement of different nucleotide compared to the reference sequence.

172

University of Ghana http://ugspace.ug.edu.gh

Figure E: Chromatogram showing SNP within beta-tubulin gene at position 1545. The first sequence at the upper part of the figure is the reference sequence and the rest of the sequences below it represent the samples. The chromatograms of the respective sequences are shown at the lower part of the figure. Each row of the chromatogram represent a sample. The different coloured peaks are the various nucleotides as follows: Green = Adenine (A), Blue = Cytosine (C), Black = Guanine (G) and red = Thymine (T). The black boxes vertically aligned indicate the SNP position (1545). Two peaks at the same position indicate heterozygosity at that locus of the gene. Where there is a complete change of nucleotide, there is a replacement of different nucleotide compared to the reference sequence.

173

University of Ghana http://ugspace.ug.edu.gh

Figure F: Chromatogram showing SNP within beta-tubulin gene at position 1555. The first sequence at the upper part of the figure is the reference sequence and the rest of the sequences below it represent the samples. The chromatograms of the respective sequences are shown at the lower part of the figure. Each row of the chromatogram represent a sample. The different coloured peaks are the various nucleotides as follows: Green = Adenine (A), Blue = Cytosine (C), Black = Guanine (G) and red = Thymine (T). The black boxes vertically aligned indicate the SNP position (1555). Two peaks at the same position indicate heterozygosity at that locus of the gene. Where there is a complete change of nucleotide, there is a replacement of different nucleotide compared to the reference sequence. .

174

University of Ghana http://ugspace.ug.edu.gh

Figure G: Chromatogram showing SNP within P-glycoprotein gene at position 1744. The first sequence at the upper part of the figure is the reference sequence and the rest of the sequences below it represent the samples. The chromatograms of the respective sequences are shown at the lower part of the figure. Each row of the chromatogram represent a sample. The different coloured peaks are the various nucleotides as follows: Green = Adenine (A), Blue = Cytosine (C), Black = Guanine (G) and red = Thymine (T). The black boxes vertically aligned indicate the SNP position (1744). Two peaks at the same position indicate heterozygosity at that locus of the gene. Where there is a complete change of nucleotide, there is a replacement of different nucleotide compared to the reference sequence.

175

University of Ghana http://ugspace.ug.edu.gh

Figure H: Chromatogram showing SNP within P-glycoprotein gene at position 2002. The first sequence at the upper part of the figure is the reference sequence and the rest of the sequences below it represent the samples. The chromatograms of the respective sequences are shown at the lower part of the figure. Each row of the chromatogram represent a sample. The different coloured peaks are the various nucleotides as follows: Green = Adenine (A), Blue = Cytosine (C), Black = Guanine (G) and red = Thymine (T). The black boxes vertically aligned indicate the SNP position (2002). Two peaks at the same position indicate heterozygosity at that locus of the gene. Where there is a complete change of nucleotide, there is a replacement of different nucleotide compared to the reference sequence.

176

University of Ghana http://ugspace.ug.edu.gh

Figure I: Chromatogram showing SNP within P-glycoprotein gene at position 5403. The first sequence at the upper part of the figure is the reference sequence and the rest of the sequences below it represent the samples. The chromatograms of the respective sequences are shown at the lower part of the figure. Each row of the chromatogram represent a sample. The different coloured peaks are the various nucleotides as follows: Green = Adenine (A), Blue = Cytosine (C), Black = Guanine (G) and red = Thymine (T). The black boxes vertically aligned indicate the SNP position (5403). Two peaks at the same position indicate heterozygosity at that locus of the gene. Where there is a complete change of nucleotide, there is a replacement of different nucleotide compared to the reference sequence.

177

University of Ghana http://ugspace.ug.edu.gh

Figure J: Chromatogram showing SNP within P-glycoprotein gene at position 5505. The first sequence at the upper part of the figure is the reference sequence and the rest of the sequences below it represent the samples. The chromatograms of the respective sequences are shown at the lower part of the figure. Each row of the chromatogram represent a sample. The different coloured peaks are the various nucleotides as follows: Green = Adenine (A), Blue = Cytosine (C), Black = Guanine (G) and red = Thymine (T). The black boxes vertically aligned indicate the SNP position (5505). Two peaks at the same position indicate heterozygosity at that locus of the gene. Where there is a complete change of nucleotide, there is a replacement of different nucleotide compared to the reference sequence.

178

University of Ghana http://ugspace.ug.edu.gh

Figure K: Chromatogram showing SNP within P-glycoprotein gene at position 5546. The first sequence at the upper part of the figure is the reference sequence and the rest of the sequences below it represent the samples. The chromatograms of the respective sequences are shown at the lower part of the figure. Each row of the chromatogram represent a sample. The different coloured peaks are the various nucleotides as follows: Green = Adenine (A), Blue = Cytosine (C), Black = Guanine (G) and red = Thymine (T). The black boxes vertically aligned indicate the SNP position (5546). Two peaks at the same position indicate heterozygosity at that locus of the gene. Where there is a complete change of nucleotide, there is a replacement of different nucleotide compared to the reference sequence.

179

University of Ghana http://ugspace.ug.edu.gh

Appendix VII: Estimation of genotype frequencies and F-statistics

A. Estimation of genotype frequencies Assuming the following genotypes are present (observed) in a population and the quantity are as follows: AA = 50 AB = 30 BB = 40

AA represents the homozygote trait of one genotype whiles BB represents the homozygote trait of another genotype. AB therefore represent the heterozygote trait after mating of AA and BB. The total number in the population from the above quantities is 120.

The observed genotype frequencies are as followings: AA = 50/120 = 0.42 AB = 30/120 = 0.25 BB = 40/120 = 0.33 NB: The frequencies could be expressed as a fraction or a percentage.

Observed allele frequency of A AA = 50 x 2 =100 AB = 30 x 1 = 30 Total alleles of A in the population = 100 + 30 = 130 Total alleles of A and B in the population = 120 x 2 = 240 Allele frequency of A = 130/240 = 0.54

Observed allele frequency of B BB = 40 x 2 =80 AB = 30 x1 = 30 Total alleles of B in the population = 80 + 30 = 110 Total alleles of A and B in the population = 120 x 2 = 240 180

University of Ghana http://ugspace.ug.edu.gh

Allele frequency of B = 110/240 = 0.46

B. Estimation of expected genotype frequency using Hardy-Weinberg Equation (HWE) Based on the observed allele frequencies estimated above, the expected genotype frequencies using HWE are estimated as follows: Genotypes = AA AB BB HWE = p2 + 2pq + q2 = 1

Where p = allele frequency of A q = allele frequency of B

Expected AA frequency = (0.54)2 = 0.29 Expected AB frequency = 2 (0.54 x 0.46) = 0.50 Expected BB frequency = (0.46)2 = 0.21

C. Testing the difference between observed and expected (HWE) genotype frequencies:

Genotypes = AA AB BB HWE = p2 + 2pq + q2 = 1

Chi2 = (O – E)2/E + (O – E)2/E + (O – E)2/E Where O = observed genotype frequency E = expected genotype frequency Chi2 = (0.42-0.29)2/0.29 + (0.25-0.50)2/0.50 + (0.33-0.21)2/0.21 = 0.06 + 0.13 + 0.07 Chi2 = 0.26

NB: Use the Chi2 table with one degree of freedom and the above Chi2 to estimate the P value. The P value indicate whether the observed genotype frequencies are statistically significantly different from the expected genotype frequencies using HWE. With the above example, P value = 0.61 (not significant).

181

University of Ghana http://ugspace.ug.edu.gh

D. Inbreeding Coefficient (F-statistic) estimation Hardy Weinberg Equation (HWE) is a central concept that is used in the derivation of F- statistics. In a random mating population the genotype frequencies observed from generation to generation does not differ significantly from the predicted frequencies (using HWE) derived using the observed allele frequencies. Where there is no random mating, a sub- population may be created which results in a reduction in heterozygosity than observed. Based on the above observed genotype and allele frequencies, the inbreeding coefficient of the population is estimated below:

Genotypes = AA AB BB

HWE = p2 + 2pq + q2 = 1 NB: Expected Homozygosity = p2 + q2 Expected Heterozygosity = 2pq Expected Heterozygosity = 1 – Homozygosity Expected Heterozygosity = 1 – (p2 + q2)

Where p = allele frequency of A q = allele frequency of B

Using the allele frequencies of A and B estimated above:

The inbreeding coefficient is given by FIS = (Expected Heterozygosity – Observed Heterozygosity)/Expected Heterozygosity Expected Heterozygosity (AB) = 2 (0.54 x 0.46) = 0.50 Observed Heterozygosity (AB) = 30/120 = 0.25

FIS = (0.50 – 0.25)/0.50 = 0.5 Positive inbreeding coefficient indicate a loss of heterozygosity (in-breeding) due to non- random mating within sub-populations and negative values indicate a gain in heterozygosity (out-breeding).

182

University of Ghana http://ugspace.ug.edu.gh

Appendix VIII: Publication (Clinical Infectious Disease Journal)

183

University of Ghana http://ugspace.ug.edu.gh

184

University of Ghana http://ugspace.ug.edu.gh

185

University of Ghana http://ugspace.ug.edu.gh

186

University of Ghana http://ugspace.ug.edu.gh

187

University of Ghana http://ugspace.ug.edu.gh

188