Pharmacovigilance of mass drug administration as preventive chemotherapy to control and eliminate lymphatic filariasis in Tanzania

Abstract

Mass drug administration (MDA) campaigns are usually conducted on an annual basis by many countries globally where neglected tropical diseases (NTDs), including lymphatic filariasis (LF), are endemic. Such campaigns are organized and executed by the Neglected Tropical Diseases Control Programmes (NTDCP) or the National Programmes for Elimination of LF (NPELF) whichever is applicable. During such campaigns, drugs are normally distributed randomly to at-risk populations to halt the transmission of diseases. It has been proven through research and evidence gathered through the World Health Organization (WHO) and NTDCP that such a move helps to stop the spreading of LF in affected communities. Nevertheless, in these campaigns, individuals subjected to such preventive chemotherapy are in all cases not tested for LF diagnosis before drug administration. Depending on disease prevalence, elimination status, distribution coverage, and availability of resources, such campaigns may be conducted once or twice a year.

During or after these campaigns evidence shows that, the concept of pharmacovigilance is not being considered in most of the cases and that drugs are distributed randomly without collecting safety information to be certain that the same are not causing any harm to exposed individuals. Pharmacovigilance (i.e., the science and activities relating to the detection, assessment, understanding, and prevention of adverse effects or any other medicine or vaccine-related problem) is still weak in many resource-constrained countries. Such countries are to-date grappling to set up systems that would allow for the effective collection of data on drug-related adverse events (AEs) and reactions (ADRs). Underreporting is one of the mounting challenges and National Medicines Regulatory Authorities (NMRAs) do not have adequate resources or machinery to counter the status quo. Long-term and chronic undesirable events due to prolonged use of MDA drugs including those affecting the liver and kidneys are even of more concern.

In this PhD programme, we began by conducting a study to determine the prevalence and correlates of LF infection in the Mkinga district which is located in the Tanga region on the offshores of the Indian Ocean - the northeastern part of Tanzania (Paper I). The rationale of conducting such a study was to find out the level of antigenemia and microfilaremia as the same have an impact on the safety profile of drugs used for NTDs including LF (i.e., ivermectin and albendazole – IA). In this cross-sectional community-based survey, a total of 4115 individuals (49.7% males, 35.2% children) were screened for circulating filarial antigens (CFA), microfilaremia (mf), and disease manifestations in 15 villages between November 2018 and January 2019. MDA uptake in the previous year (2017) was also assessed. The overall prevalence of CFA-positivity was 5.8% (239/4115; 95% CI: 5.1–6.6), with significant heterogeneity seen between villages (range 1.2% to 13.5%). CFApositivity was higher in males (8.8%) than females (3.3%) and correlated with increasing age (p < 0.001). Prevalence of mf amongst CFA-positives was 5.2%. The MDA uptake was only 60%. Prevalence of scrotal enlargement, hydrocele, swelling of arms or legs, lymphoedema, and lymphadenopathy was 6.4%, 3.7%, 1.35%, 1.2% and 0.32%, respectively. This study showed that the elimination target of <1% mf and <2% antigenemia to levels where recrudescence is unlikely to occur was not attained as compared to the baseline data. However, after 16 rounds of MDA, LF transmission was significantly reduced.

Following the prevalence study, we went on to conduct the safety surveillance study (Paper II) to document AEs that occurred following IA MDA. In this study, we identified the type, incidence, and associated risk factors. Around 9,640 eligible individuals received single-dose IA combination preventive chemotherapy and treatment-associated AEs were actively monitored through house-to-house visits on day 1, day 2, and day 7 of MDA. After MDA, 9288 participants (96.3%) were followed up of whom 442 reported 719 MDA-associated AEs. The incidence of experiencing one or more types of MDA-associated AE was 4.8% (95% CI = 4.3– 5.2%); this was significantly higher among those with pre-MDA clinical events than those without (8.5% versus 4.1%, p < 0.001). AEs were mild (83.8%), moderate (15.9%), and severe (0.3%), and most resolved within 72 hours. The incidence of experiencing one, two, and ≥ three types of AEs were 2.8%, 1.3%, and 0.6%, respectively. The most common AEs were headache (1.23%), drowsiness (1.15%), fever (1.12%), and dizziness (1.06%). Chronic illness, clinical manifestation of LF, being female, or having pre-existing clinical symptoms were significant predictors of AEs. Therefore, safety monitoring in individuals with underlying clinical conditions was recommended for timely detection and management of AEs during MDA campaigns.

To further ascertain the safety of MDA drugs in blood, kidneys, and liver, we extracted data to measure the haematological and biochemical parameters to determine their changing patterns (Paper III). In so doing we also examined their relationship with those who had experienced AEs in the previous safety study. In this nested analytical prospective study, we assessed data amongst 499 eligible individuals whose blood samples were collected before and after MDA. We measured Complete Blood Count (CBC) and renal and liver function tests at days 0 and 7 following MDA. After a comprehensive assessment of all blood indices as well as kidney and liver surrogate markers, we demonstrated that haematological and biochemical changes may occur following IA MDA. The median values of haematological parameters, including RBC, Hb, and HCT decreased while MCH, MCHC, and monocytes increased significantly in both CFA-positive and negative individuals (p<0.05). Median platelet counts, including MPV and PLCR, increased after MDA (p<0.05). Biochemical parameters, including ALT, AST, BilD, and BildT, showed varied results. Higher creatinine levels were observed, and a significant proportion of individuals had haematological and biochemical parameters below the reference range. Drowsiness was the most observed AE among individuals with abnormal parameters, followed by fever, dizziness, and nausea.

In Paper IV, a pharmacokinetic (PK) study was conducted to determine the PK properties of IVM in humans. The study was done to characterize the disposition of IVM and determine predictors of its PK for dose optimization during MDA. This was also a nested study in a bigger safety study in which data was evaluated amongst 468 individuals. PK samples were collected at 0, 2, 4, and 6 hours from individuals weighing > 15 Kg receiving IVM (3-, 6-, 9-, or 12 mg) and ALB (400 mg) during an MDA campaign. Individual characteristics were assessed including demographics, laboratory/clinical parameters, and genetic variations of selected drug-metabolizing enzymes and transporters. IVM plasma concentrations were quantified by LC-MS/MS and analysed using population-pharmacokinetic (POPPK) modelling. A two-compartment model with transit absorption kinetics, and allometrically scaled oral clearance (CL/F) and central volume (Vc/F) were adopted. Fitting of the model to the current data identified a 48% higher bioavailability for the 3 mg dose compared to other doses and further identified a subpopulation with 97% higher mean transit time (MTT). The final estimates for CL/F, Vc/F, inter-compartment clearance (Q), peripheral volume (Vp), MTT, and absorption rate constant (Ka) for a 70 Kg person (on a dose other than 3 mg) were 7.7 L/h, 147 L, 20.4 L/h, 207 L, 1.5 h, and 0.71/h, respectively. Simulations indicated that weight-based dosing provides comparable exposure across weight bands, but height-based dosing with a capping IVM dose at 12 mg for individuals with height > 160cm under-doses those weighing > 70 Kg. The variability in IVM PK is partly explained by body weight and dose. The established POPPK model can be used for IVM dose optimization. Height-based pole dosing may result in varying IVM exposure of individuals in different weight bands; hence use of weighing scales for IVM dosing during MDA is recommended.

In Paper V, we investigated the efficacy of IA in clearing mf and reducing CFA levels in individuals. This community-based prospective study assessed the efficacy of MDA drugs in mf clearance and CFA reduction on days 7- and 6 months following MDA. The study was done in the same Mkinga district, Tanga region between November 2018 and June 2019. The status of mf and CFA on day 7 and six-month post-MDA was monitored. The primary efficacy outcomes were the clearance rates of mf on day 7 and six months, and CFA at 6 months of post-MDA. Out of 4,115 individuals screened, 239 (5.8%) tested positive for CFA, with 11 (4.6%) also positive for mf. The McNemar test revealed a significant improvement in mf clearance on day 7 following MDA (p=0.02). Out of 183 CFA-positive individuals who were available at 6-month follow-up, 160 (87.4%) remained CFA positive, while 23 became CFA negative. The CFA clearance rate at 6 months post-MDA was 12.6% (95% CI = 8.52 – 18.5%). There was no significant association of variability in IVM plasma exposure (Cmax and AUC) with post-MDA mf or CFA clearance status. We concluded that preventive chemotherapy with IA effectively clears mf within a week, but the same drugs are ineffective in clearing CFA at six months post-MDA. We recommended alternative drug combinations targeting adult worms as IA is not macrofilaricidal.

In conclusion, LF has not been fully eliminated from the study area and the prevalence of the disease is still above the threshold recommended by WHO. MDA drugs are relatively safe for use during MDA. However, chronic illnesses, previous clinical manifestations of LF, being female, or having pre-existing clinical symptoms are significant predictors of AEs. Individuals exposed to MDA might experience changes in haematological and biochemical parameters after drug intake. The NTD programme should consider these findings when monitoring individuals taking part in future MDA campaigns. POPPK modelling can be used for IVM dose optimization and whenever possible, the programme should consider using weighing balances rather than height-poles for dose estimation when administering MDA drugs to individuals in endemic communities. Despite that the MDA drugs are still efficacious, alternative combinations targeting adult worms should be sought by the NTD programme.