Ecological risk modelling of antiretroviral drugs in the aquatic environment

Abstract

The success of antiretroviral therapy (ART) programme's on the management of the Human Immunodeficiency Virus/ Acquired Immunodeficiency Syndrome (HIV/AIDS) in the Sub-Saharan African (SSA) region has inadvertently led to the release of antiretroviral drugs (ARVs) into the environment. This study aimed to model the quantities and risks of ARVs in aquatic environments. Herein, the ecological risks of first regime (efavirenz (EFV), lamivudine (3TC), nevirapine (NVP), zidovudine (AZT) and tenofovir (TDF)), and second regime ARVs (lopinavir (LPV), atazanavir (ATV), ritonavir (RTV), didanosine (DDI) and dolutegravir (DTG)) were investigated. ARV releases and quantification into the aquatic environment were based on Eswatini as a study case country using the material flow analysis (MFA) approach. Quantities of ARVs into the environment were estimated for all ten drugs using the input model parameters, viz.: (number of people living with HIV (PLWHIV), number of people on ARVs, drug dosages, treatment adherence and drug excretion rate). The predicted environmental concentrations (PEC) were then compared to the predicted no effect concentrations (PNEC) to estimate the possible risk (by deriving the risk quotient (RQ). Owing to the dearth of ecotoxicological data for ARVs, Ecological Structure Activity Relationships (ECOSAR) was used to derive the toxicity of each drug for three species: daphnia, algae, and fish. Results indicated quantities of ARVs were three-fold higher in surface water relative to wastewater. For example, in 2017, EFV was 5 671 and 1 853 kg in surface water and wastewater, respectively. Higher flows of ARVs in surface waters were associated with a lack of sanitation services for most Eswatini population, and low wastewater treatment plant (WWTP) performance. Results further showed wide variability of PECs in surface water (0.08–73.2 μg/L) and wastewater (0.02 – 23.1 μg/L). Of the ten ARVs, the risk of DTG was unquantifiable due to a lack of ecotoxicity data even when using the ECOSAR approach. Only DDI had RQs < 1 for all three trophic levels in surface water, whereas EFV, LPV, and RTV posed the highest risk (RQ >10) to all three taxa. Further, NVP and AZT presented moderate risk (1 < RQ < 10) to algae, and 3TC and TDF posed minimal risk (RQ < 1) to daphnia. No risk for the ARVs was estimated in wastewater except for TDF due to the lack of their bacterial toxicity data. Results herein make it possible for the first time to quantify the potential environmental risks of ARVs using a modelling approach. The model can be extended to other SSA countries given their limited capabilities to detect and quantify ARVs risks in the environment using costly analytical techniques. Furthermore, the results suggest lack of sanitation services and high HIV/AIDs prevalence in SSA with concomitant high concentrations of ARVs in the environment potentially may trigger antiretroviral drug resistance, with possible negative implications to the ART programme.