Between 2000 and 2019, malaria-related deaths worldwide were reduced by 44%, from 736,000 in 2000 to 409,000 in 2019 (World Health Organization, 2020). Nevertheless, the spread of the COVID-19 pandemic to all malaria-endemic countries was responsible for the disruption of essential malaria services (World Health Organization, 2020). This has compromised the WHO global strategy, which had been designed to reduce the incidence of malaria and mortality rates by at least 90% by 2030 (World Health Organization, 2015). In 2020, Brazil reported approximately 141,000 cases of malaria (84% caused by Plasmodium vivax and 16% by Plasmodium falciparum) (BRASIL. Ministério da Saúde, 2021). Although these numbers represent a reduction in the total number of cases of P. vivax, there was a 32.6% increase in P. falciparum cases when compared with 2019.

Early diagnosis and treatment are fundamental components of strategies designed to control and eliminate malaria. In many endemic areas, the routine diagnosis is primarily performed microscopically. Nonetheless, the rapid diagnostic tests (RDTs) provide a useful tool when microscopy facilities are unavailable, particularly in remote areas and in the absence of a well-trained professionals (BRASIL. Ministério da Saúde, 2009). RDTs can be used to detect specific Plasmodium antigens in the blood stream, such as histidine-rich protein 2 (HRP2) and lactate dehydrogenase (LDH) (World Health Organization, 2016). The sensitivity of RDTs is dependent on detection of the target protein; however, since the 2000s, isolates with deletion of the gene pfhrp2 gene have been reported in Africa and South America (Baker et al., 2005; Baker et al., 2010; Gamboa et al., 2010; Houzé et al., 2011). In Brazil, pfhrp2 deletions have been reported in some endemic states, such as Acre, Rondônia, and Amazonas (Rachid Viana et al., 2017; Góes et al., 2021).

PfHRP2 comprises different types of histidine and alanine repeats (Rock et al., 1987). An extensive study conducted by the Malaria RDT Quality Assurance Program reported a high variability in the pfhrp2 sequence of isolates from Asia and the southwest Pacific (Baker et al., 2005; Lee et al., 2006), and it was subsequently proposed that certain types of alanine/histidine repeats were associated with variations in the sensitivity of RDTs, interfering with the test efficacy (Baker et al., 2005; Lee et al., 2006). However, other studies have shown contrasting data concerning the effect of PfHRP2 variability on RDT efficacy (Baker et al., 2010; Kumar et al., 2012; Atroosh et al., 2015; Willie et al., 2018).

In recent years, reports of pfhrp2 deletions have increased, including 16 reports from 15 endemic countries only between 2019 and 2020 (World Health Organization, 2020). The WHO is currently tracking these publications to monitor pfhrp2 deletions and recommends that countries conduct representative baseline surveys among suspected malaria cases with false-negative RDT results (World Health Organization, 2020). In this study, we evaluated the proportion of pfhrp2 deletion in P. falciparum isolates from four malaria-endemic states of the Brazilian Amazon (Amapá, Mato Grosso, Rondônia and Roraima state) over a period of 18 years (2002 – 2020). RDT results were interpreted considering the pfhrp2 profile, parasitemia, and complexity of infection. Additionally, a comprehensive analysis of PfHRP2 sequence variability was conducted for parasite isolates from South America.

Deletion of pfhrp2/3 in the P. falciparum genome raises concerns regarding the efficacy of RDTs, given that parasites characterized by such deletions may not be identified by RDTs that are based on the detection of histidine-rich protein 2 (HRP2). According to WHO guidelines, the threshold for RDT change is a pfhrp2 deletion rate greater than 5% causing false-negative RDT results (World Health Organization, 2020). In the present study, we assessed the pfhrp2 deletion profile of 132 samples from four different states of the Brazilian Amazon. Our findings revealed that although the deletion of this gene is present in these regions, the proportion of false-negative RDT results attributable to pfhrp2 deletion appears to be very low (1%). These findings can be attributed to polyclonal infections, owing to the presence of genetically distinct parasites differing in PfHRP2 expression that infect the same host. However, at this time we cannot exclude the possibility of cross-reactivity with PfHRP3.

Overall, 10% of the P. falciparum isolates analyzed were found to have a pfhrp2 deletion, which was confirmed by the finding that most isolates were also frequently characterized by deletions of the genes flanking pfhrp2. Isolates in samples collected from Roraima were found to have the highest proportion (15%) of pfhrp2 gene deletion, followed by the Mato Grosso (10%), and Rondônia (6%) isolates. These findings contrast with previous observations in other areas of the Brazilian Amazon, with rates of 32% to 72% and 54% to 100% being reported in the states of Acre and Amazonas, respectively (Rachid Viana et al., 2017; Góes et al., 2021). More consistent with our present findings, Rachid Viana et al. (2017) reported a low proportion (3%) of pfhrp2 deletions in Rondônia state for isolates sampled in the Monte Negro municipality. We and others have also found that the MAL7P1_230 gene flanking pfhrp2, which encodes a Plasmodium exported protein (PHIST) of unknown function (https://plasmodb.org/), is frequently deleted in Brazilian samples (range = 21%–45%) (Rachid Viana et al., 2017), whereas the other flanking gene, MAL7P1_228, encoding for a putative heat shock 70 protein, was less frequently deleted (range = 0.5%–11%) in the same samples.

The present study is the first to assess the deletion profile of pfhrp2 in isolates from Amapá state, and although we were only able to analyze a relatively few samples, none of these showed deletions of this gene. Consistently, previous studies in the neighboring Brazilian state of Pará and French Guiana have reported the absence of pfhrp2 deletions in isolates obtained from both areas (Trouvay et al., 2013; Rachid Viana et al., 2017). In contrast, as previously mentioned, we detected a high proportion of pfhrp2-negative isolates (15%) from Roraima state, which borders Guyana and Venezuela, and notably, P. falciparum samples collected in Roraima were mainly obtained from travelers or workers originating from Venezuela, a country that has contributed significantly to an increase in imported cases of malaria in this region (BRASIL. Ministério da Saúde, 2020; Arisco et al., 2021). Our findings accordingly underscore the need for constant surveillance, particularly in this area, which has been established to have the highest number of cross-border malaria cases in Brazil.

Despite the importance of monitoring for the presence of parasites with a pfhrp2 deletion, we showed that the data obtained requires careful analysis. By applying two well-established PCR-based protocols, we found discrepancies in the results obtained regarding the proportions of pfhrp2 deletion in the same samples. Specifically, whereas the protocol based on conventional PCR indicated a 46% deletion, that based on nested PCR indicated a rate of only 10%. In this context, Parr et al. (2018) recently addressed issues concerning the sensitivity and specificity of PCR assays for pfhrp2. The authors demonstrated a tenfold improvement in the limit of detection when the extension temperature of the PCR protocol was lowered from 72°C to 60°C (Parr et al., 2018). In the present study, based on an analysis of three polymorphic microsatellite locus, we showed that most infections were characterized by multiple clones of genetically distinct variants of the parasite. We accordingly speculate that such polyclonal infections may complicate analyses using PCR-based testing methods, given that different parasite populations with and without pfhrp2 may be present in the same infected individual. For example, polyclonal infections can confound reporting of the true proportion of gene deletions, whereas RDT results may be influenced by the presence of genetically distinct parasites with varying degrees of PfHRP2 expression in the same infection. We detected eight samples with pfhrp2 deletions and positive results in the RDT; among these, all but one sample harbored polyclonal infections. Preliminary data based on conventional PCR indicated that these samples were pfhrp3-negative. However, a method of greater sensitivity needs to be employed to enable a more definitive conclusion regarding pfhrp3 deletion. Thus, even though this result could be explained in terms of the presence of pfhrp3, the diversity of P. falciparum populations within the infections could also result in parasite detection by RDTs.

The samples analyzed in this study were collected over a period of 18 years (2002 – 2020), and thus the results obtained provide an indication of how the pfhrp2 deletion profile has evolved over the years in the Brazilian Amazon. During the assessed period, we detected a moderate and relatively consistent proportion (10%) of pfhrp2 deletions. In contrast, differing rates of pfhrp2 deletion have previously been reported in Peru and Colombia. For example, Dorado et al. (2016) recorded a deletion rate of 6% in samples from Colombia collected from different sites between 2003 and 2010, whereas rates reaching up to 18% were detected in samples collected from 1999 to 2009 in a study conducted by Murillo Solano et al. (2015) (Dorado et al., 2012; Murillo Solano et al., 2015; Dorado et al., 2016). Similarly, in Peru, two studies reported distinct deletion rates for samples collected at different sites approximately 5-years apart, with Gamboa et al. (2010) reporting a 41% of deletion (samples collected between 2003 and 2007), whereas Maltha et al. (2012) detected 26% pfhrp2-negative isolates (2010 and 2011). Continuous surveillance will provide a clearer picture of how the proportions of gene deletions change over time and across regions.

Although the sequences of pfhrp2 tend to show considerable diversity worldwide, this variability does not appear to have a substantial influence on the performance of RDTs (Baker et al., 2010; Kumar et al., 2012; Atroosh et al., 2015; Willie et al., 2018). There is, however, evidence to indicate that polymorphisms in the amino acid sequence of PfHRP2 may affect RDT sensitivity at very low parasite densities (<200 parasites/μL) (Baker et al., 2010). To better characterize the variability of PfHRP2, we sequenced the genes of P. falciparum isolates representative of the four regions assessed in the present study. As has previously been reported for isolates from other countries (Baker et al., 2005; Baker et al., 2010), we found that all sequences started with a type 1 repeat, ended with a type 12 repeat, and comprised mainly type 2, 6, and 7 repeats. Interestingly, in this regard, a significant relationship has been described between the number of type 2 and 7 repeats and RDT sensitivity at low parasitemia (Baker et al., 2005; Kumar et al., 2012). Nonetheless, there is a yet no consensus regarding this association, and it remains to be established how PfHRP2 variability affects RDT performance (Baker et al., 2010; Kumar Bharti et al., 2017).

We identified 10 distinct patterns of the PfHRP2 sequence based on the type and number of histidine/alanine repeat units present in each sequence, which were mostly exclusive of each region. The only exception in this regard was the pattern III, which was observed at a high frequency in isolates from Rondônia and Mato Grosso collected during the same period. Notably, however, many samples collected in Mato Grosso were from individuals who were probably infected in Rondônia. The high genetic diversity of the parasite population from Roraima was associated to a high variability of PfHRP2. In contrast, for samples collected in the other states, there was not a straightforward association between levels of diversity of parasite and PfHRP2. A comparison of the Brazilian PfHRP2 sequences obtained in the present study with those identified in other South American countries revealed that sequences with similar profiles have been detected in Brazil and French Guiana, particularly the three patterns (I, V, and VI) found predominantly in Roraima and Amapá states. The northeast borders of Brazil that include these two states comprise the “Guiana shield”, along with Suriname, Guyana, French Guiana, and parts of Colombia and Venezuela. This region is characterized by large-scale movements of illegal gold miners and a large number of cross-border malaria cases (Ferreira and Castro, 2019; Arisco et al., 2021). Hence, we might expect to detect extensive gene flow and common patterns of PfHRP2 sequences.

In conclusion, we detected low to moderate levels of pfhrp2 deletion in different regions of the Brazilian Amazon. Importantly, the presence of pfhrp2-negative isolates has not been translated into a reduction in the efficacy of RDTs. We believe that these observations can be attributed, at least in part, to the multiplicity of infection in individuals; although at present, we cannot exclude the possibility of cross-reaction with PfHRP3. We have focused on describing the variability of pfhrp2 in the Brazilian Amazon, given that PfHRP2 is the main antigen targeted by most RDTs. However, there is evidence to indicate that some monoclonal antibodies used in PfHRP2-based RDTs show cross-reactivity to PfHRP2 and PfHRP3 (Lee et al., 2006). In Brazil, similar to the findings reported from other South American countries (40%–70% of pfhrp3 deletion), the proportion of pfhrp3 deletions is typically very high (35%–98%) (Gamboa et al., 2010; Murillo Solano et al., 2015; Dorado et al., 2016; Rachid Viana et al., 2017; Góes et al., 2021). One of the limitations of the present study is that we were unable to assess PfHRP2 levels or the performance of a second brand of RDT to confirm initial results indicating a lack of the protein. Nevertheless, an important finding in this respect was the discrepancy we identified in the proportions of pfhrp2 deletion detected using two different PCR protocols, which reinforces the need to pay particular attention to laboratory workflows when seeking to assess gene deletions. Taken together, our findings highlight the importance of continuously monitoring the presence and spread of parasites with a pfrhp2 deletion in the Amazon region to ensure optimal RDT performance.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: GenBank MZ773486, MZ773487, MZ773488, MZ773489, MZ773490, MZ773491, MZ773492, MZ773493, MZ824674, MZ824675, MZ824676, MZ824677, MZ824678, MZ824679, MZ824680, MZ824681, MZ824682, MZ824683, MZ824684, MZ824685, MZ824686, MZ824687, MZ824688, MZ824689, MZ824690, MZ824691, MZ824692, MZ824693, MZ824694, MZ824695, MZ824696, MZ824697, MZ824698, MZ824699.

The ethical and methodological aspects of this study were approved by the Ethical Committee of Research on Human Beings of the René Rachou Institute (N° 2.243.058), according to the Brazilian National Council of Health (Resolutions 196/96 and 466/ 12). All adult participants signed a written informed consent, whereas next of kin, caretakers, or guardians signed on the behalf of the minors/children enrolled in the study. All the methods were carried out in accordance with the approved guidelines.

TS and CJF were the principal investigators. MM, TM, and LF performed the PCR analysis. GC and MM performed sequencing and genotyping analysis. TS, CJF, GC, JL, DP, and AA helped carry out the field work. TS, CJF, CFA, LC, DP, AA, JL, and GC participated in interpretation of the data and critical revisions of the manuscript. All authors contributed to the article and approved the submitted version.

This study was funded by the Royal Society of Tropical Medicine and Hygiene (RSTMH Small Grants 2016), Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Programa para Inserção de Recém-Doutores nos Programas de Pós-Graduação da Fiocruz and the Programa PrInt-Fiocruz-CAPES. TS, CJF, CFA, and LC are CNPq Research Productivity fellows. GLC thanks the FAPEMIG for scholarship support. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication. This study was partially supported by the Coordination for the Improvement of Higher Education Personnel (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES) - Finance Code 001.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.