Artemisinin-based combination therapies (ACTs) are the mainstay of malaria therapy. Since their introduction about 20 years ago, they have prevented the death of millions (World Health Organization, 2023). Today, resistance to artemisinin and its partner drugs is threatening the effectiveness of ACTs (World Health Organization, 2023). Indeed, reports of ACT treatment failures are mounting. Amongst the six ACTs currently recommended by the World Health Organization for the treatment of uncomplicated malaria (World Health Organization, 2022), dihydroartemisinin-piperaquine (DHA-PPQ) experiences most treatment failures. For instance, half of the therapeutic efficacy studies conducted in the WHO Western Pacific Region between 2015 and 2021 reported treatment failure in at least 10% of the participants following DHA-PPQ therapy (World Health Organization, 2023).

Whilst reduced susceptibility to DHA (or artemisinin) is well described (Teuscher et al., 2010; Witkowski et al., 2010; Witkowski et al., 2013; Ariey et al., 2014; Birnbaum et al., 2020), the mechanism of PPQ resistance is not fully elucidated yet. Elevated 50% and 90% inhibitory concentrations (IC₅₀ and IC₉₀) of PPQ immediately determined ex vivo in Cambodian patient isolates were detected as early as 2013 and showed a significant association with DHA-PPQ treatment failure (Saunders et al., 2015; Chaorattanakawee et al., 2016). In addition, culture-adapted, k13-mutant parasites isolated from DHA-PPQ-treated patients in Cambodia repeatedly demonstrated abnormal dose-response curves in in vitro growth inhibition assays following PPQ exposure (Duru et al., 2015; Bopp et al., 2018; Ross et al., 2018). These curves were characterized by persistent parasite growth, i.e. incomplete growth inhibition, or even an increase in parasite growth at high drug concentrations (~200 nM and higher), and hence were often difficult to interpret. To allow for better monitoring of PPQ resistance, Duru et al. developed the PPQ survival assay (PSA_0-3h), a washout assay tailored to the pharmacological profile of the drug in humans (Duru et al., 2015). Increased survival in the PSA_0-3h (i.e. > 10%) was not only associated with abnormal dose-response curves, but also with treatment failure in the patients from whom the parasite isolate originated (Duru et al., 2015). Finally, studies aiming at identifying genetic markers of PPQ resistance pointed to an association with amplifications in plasmepsin 2 and 3 (Agrawal et al., 2017; Amato et al., 2017; Witkowski et al., 2017; Bopp et al., 2018; Kane et al., 2023), or single nucleotide polymorphisms in exonuclease (Amato et al., 2017) and chloroquine-resistance transporter (crt) (Duru et al., 2015; Agrawal et al., 2017; Dhingra et al., 2017; Ross et al., 2018; Dhingra et al., 2019; Gomez et al., 2023; Kane et al., 2023).

Compounds exhibiting an incomplete growth inhibition phenotype are routinely being picked up in drug screening activities conducted on a panel of drug-resistant field isolates, e.g. as observed in a lead-optimization study on 2,6-imidazopyridines (Le Manach et al., 2018). In fact, this phenotype is considered a warning sign and often prompts closer examination. In this pilot study, we were interested in how a compound with an incomplete growth inhibition phenotype, i.e. PPQ, performs when measuring parasite viability rather parasite growth. Providing highly sensitive and rich pharmacodynamic data, parasite viability has proven to be a superior measure of drug activity than conventional readouts such as in vitro growth inhibition or in vivo parasite clearance (Sanz et al., 2012; Rebelo et al., 2020; Walz et al., 2023). Therefore, we tested PPQ in the in vitro parasite reduction ratio (PRR) assay version 2 (Walz et al., 2023), an assay used to measure parasite viability, against the drug-sensitive P. falciparum strain NF54 and the Cambodian field isolate RF12 (also known in the literature as PH1263-C (Ross et al., 2018)). We show that measuring parasite viability provides a much clearer picture of the resistance phenotype to PPQ, and hence might also have the potential to support decision-making with respect to candidate prioritization in preclinical development and treatment policy changes in the post-marketing phase.

We first wanted to reproduce the abnormal dose-response curves of PPQ published by Ross et al. (2018). For this, we ran growth inhibition assays with the pan-sensitive P. falciparum isolate NF54 and the PPQ-resistant Cambodian field isolate RF12, which carries mutations in kelch13 (C580Y) and Pfcrt (H97Y) and a single copy of Pfmdr1. Both lines were tested in n = 6 biological replicates (each with n = 2 technical replicates). Whilst the sensitive parasite isolate displayed an exemplary sigmoidal dose-response curve with a single-digit nanomolar IC₅₀ value (7.7 nM) for PPQ ( Table 1 ), the Cambodian isolate presented an abnormal dose-response phenotype that was characterized by a plateau at around 15.3% parasite growth towards the highest PPQ concentrations tested, i.e. an incomplete growth inhibition phenotype ( Figure 1A ). The median IC₅₀ of the Cambodian isolate increased by a factor of 2.5 to 19.3 nM compared to the drug-sensitive parasite isolate ( Table 1 ).

In a next step, we used the same two parasite isolates to initiate independent PRR assays in order to assess parasite viability following PPQ exposure. In contrast to the growth inhibition assay, here the parasites are serially diluted and re-cultivated in the absence of drug after a certain duration of drug exposure, so that they – if viable – can reestablish a parasite culture. In line with previous experiments conducted in-house with the drug-sensitive P. falciparum isolate NF54, PPQ treatment resulted in a fast, chloroquine-like killing profile. An immediate onset of drug activity was accompanied by a PRR of 4.4 rendering almost all parasites non-viable after 48 hours of drug exposure ( Figure 1B , Table 1 ). In contrast, when assessed in the Cambodian isolate RF12, PPQ treatment did not result in a decline of viable parasites ( Figure 1B ). Instead, the number of viable parasites increased by a factor of 10, approximately, within the first 96 hours of treatment, hence indicating a lack of drug activity in this parasite isolate.

In the recent years, field isolates of P. falciparum have repeatedly demonstrated abnormal dose-response curves following PPQ exposure in vitro. Stagnation or even increase of parasite growth towards higher PPQ concentrations was associated with increased survival in the PSA (Duru et al., 2015; Ross et al., 2018), specific genotypes (Bopp et al., 2018; Ross et al., 2018), and with increased treatment failure in DHA-PPQ-treated patients (Duru et al., 2015). With the goal to better understand this incomplete growth inhibition phenotype, we assessed the viability of a culture-adapted field isolate in the PRR assay, i.e. a viability assay. To the best of our knowledge, this is the first time a P. falciparum strain other than the drug-sensitive ones (P. falciparum 3D7 or NF54) was tested in the in vitro PRR assay. Using PPQ as an example, we found that incomplete growth inhibition translates in a drastic increase in parasite viability when assessed in a PPQ-resistant Cambodian field isolate as compared to its drug-sensitive counterpart. Importantly, PPQ was previously considered fast-acting according to the classical PRR assay conducted with the drug-sensitive P. falciparum isolate NF54 and its clone 3D7 (see Figure 1B , and Sanz et al. (2012)). In the Cambodian field isolate RF12, in contrast, the compound was classified as slow-acting or even inactive.

In that regard, it was particularly striking that the potency shift (i.e. IC₅₀ shift between drug-sensitive and -resistant parasites) was well below the threshold for drug resistance (defined as a shift ≥ 5 (Duffey et al., 2021) or even ≥ 20 (Ding et al., 2012)), indicating that a shift in IC₅₀ is not sufficient to evaluate drug susceptibility. Growth inhibition assays conducted with PPQ-resistant parasites in prior studies did also not result in a substantial IC₅₀ shift (Leang et al., 2013; Lim et al., 2013; Saunders et al., 2014; Bopp et al., 2018). Yet, the plateau towards higher drug concentrations in dose-response curves might be a first indicator of reduced drug susceptibility. Whether this holds true beyond PPQ remains to be validated in follow-up studies with additional field isolates and chemotypes. Still, minor plateaus may be easily overlooked, especially when the background signal of the underlying readout method is high. Therefore, we argue that viability assays with culture-adapted, drug-resistant field isolates may serve as complementary tool to evaluate the potency of selected, advanced candidates.

The classical PRR assay provides relevant pharmacodynamic parameters, such as the PRR or the E_max , that shed light on the in vitro killing kinetics of a compound in drug-sensitive parasites; these can be used for downstream PK/PD analysis (Wicha et al., 2022; Walz et al., 2023), e.g. to predict the effective dose of single drugs or drug combinations in humans. However, our data suggest that solely relying on pharmacodynamic data from a single, drug-sensitive parasite strain may be insufficient, potentially resulting in overestimation of the drug effect in the field setting. Instead, we recommend to assess the pharmacodynamics of advanced drug candidates in the context of an in vitro PRR assay on a representative panel of drug-resistant field isolates, similar to the one deployed for cross-resistance testing in growth inhibition assays (Chugh et al., 2015; Duffey et al., 2021). This might result in more accurate predictions of drug efficacy in actual malaria patients, ultimately reducing the risk of late withdrawal of drug candidates (and concomitant financial losses and ethical concerns), and hence streamlining the drug development process.

In vitro washout assays other than the PRR assay, such as the ring-stage survival assay (RSA_0-3h) for artemisinin derivatives (Witkowski et al., 2013; Walz et al., 2019) and the PPQ survival assay (PSA) (Duru et al., 2015), are valuable tools to measure the level of drug resistance in a short amount of time. However, their use is tailored to a specific drug, i.e. to mimic the pharmacokinetic profile (drug exposure and concentration observed in patients), with the intention to monitor the antimalarial efficacy of drugs that are already approved and routinely deployed in the field. The in vitro PRR assay, in contrast, allows predicting the risk of drug resistance of diverse preclinical candidates irrespective of their chemotype and pharmacokinetic profile.

The usefulness of parasite viability assays to assess drug resistance in vivo was also probed before. In contrast to the conventional measure of in vivo drug activity (i.e. parasite clearance (White, 2011)), parasite viability differentiates between viable and nonviable parasite populations in a sample (Rebelo et al., 2020; Radohery et al., 2022a; Radohery et al., 2022b). Not doing so might result in underestimation of drug efficacy as demonstrated for artesunate in studies conducted in mice and humans (Rebelo et al., 2020; Radohery et al., 2022b). Since in vivo viability assays provide a more sensitive estimate of drug activity than parasite clearance, they are expected to spot signs of drug resistance earlier. Using a mathematical model, Hastings et al. found that in vivo parasite clearance is highly insensitive at detecting drug resistance and does so only if resistance is sufficiently strong, as observed with kelch13 mutations (Hastings et al., 2015). Later, Rebelo et al. corroborated this theory by showing that volunteers infected with artemisinin-resistant parasites had a 2-fold longer parasite clearance half-life than volunteers infected with artemisinin-sensitive parasites, whereas the difference in half-lives was 12-fold when parasite viability was measured instead (Rebelo et al., 2020).

In conclusion, assessing parasite viability following treatment in culture-adapted field isolates has the potential to improve our understanding of antimalarial drug resistance. It will likely result in more accurate predictions of drug efficacy, thereby guiding resistance monitoring in the field and aiding decision-making in preclinical and clinical drug development. In addition, measuring parasite viability in a representative panel of field isolates might help to detect drug resistance earlier, and therefore would allow faster switching of national treatment policies, ultimately preventing the spread of drug resistance.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The manuscript presents research on animals that do not require ethical approval for their study.

AW: Conceptualization, Formal analysis, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. SS: Investigation, Validation, Writing – review & editing. CS: Investigation, Validation, Writing – review & editing. BT: Formal analysis, Visualization, Writing – review & editing. PM: Resources, Supervision, Writing – review & editing. SW: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – review & editing.