Intensity of pyrethroid resistance in An. gambiaes .l.before and after a mass distribution of insecticide treated nets in Kinshasa and from 11 provinces of the Democratic Republic of Congo UNDER REVIEW

Between 2011 and 2018, an estimated 134.8 million pyrethroid-treatedlong-lasting insecticidal nets (LLINs) were distributed nationwide in the Democratic Republic of Congo (DRC) for malaria control. Pyrethroid resistance has developed in DRC in recent years, but the intensity ofresistance and impact on LLIN efficacy was not known. Therefore, the intensity of resistance of Anopheles gambiae s.l. to permethrin and deltamethrin was monitored before and after a mass distribution of LLINs in Kinshasa in December 2016, and in 7sites across the country in 2017 and 11 sites in 2018.


Background
Between 2011 and 2018, an estimated 134.8 million pyrethroid-treatedlong-lasting insecticidal nets (LLINs) were distributed nationwide in the Democratic Republic of Congo (DRC) for malaria control. Pyrethroid resistance has developed in DRC in recent years, but the intensity ofresistance and impact on LLIN efficacy was not known. Therefore, the intensity of resistance of Anopheles gambiae s.l. to permethrin and deltamethrin was monitored before and after a mass distribution of LLINs in Kinshasa in December 2016, and in 7sites across the country in 2017 and 11 sites in 2018.

Materials and Methods
In Kinshasa, CDC bottle bioassays using 1, 2, 5, and 10 times the diagnostic dose of permethrin and deltamethrin were conducted using An. gambiae s.l. collected as larvae and reared to adults. Bioassays were conducted in four sites in Kinshasa province six months before a mass distribution of deltamethrin-treated ITNs and then two, six, and ten months after the distribution. One site in neighboring Kongo Central province was used as a control (no mass campaign of ITN distribution during the study). Nationwide intensity assays were conducted in six sites in 2017 using CDC bottle bioassays and in 11 sites in 2018 using WHO intensity assays. A sub-sample of An. gambiae s.l. were tested by PCR to determine species composition and frequency of kdr-1014F and 1014S alleles.

Results
In June 2016, before LLIN distribution, permethrin resistance intensity was high in Kinshasa; the mean mortality rate was 43% at the 5× concentration and 73% at the 10× concentration. Bioassays at 3 time points after LLIN distribution showed considerable variation by site and time and there was no consistent evidence for an increase in pyrethroid resistance intensity compared to the neighboring control site. Tests of An. gambiaes.l. in 7 sites across the country in 2017 and 11 sites in 2018 showed all populations were resistant to the diagnostic doses of 3 pyrethroids.In 2018, the intensity of resistance varied by site, but was generally moderate for all three pyrethroids, with survivors at ×5 the diagnostic dose. An. gambiae s.s. was the most common species identified across 11 sites in DRC. However in Kinshasa, An. gambiae s.s.(91%) and An. coluzzii (8%) were sympatric.

Conclusions
Moderate or high intensity pyrethroid resistance was detected nationwide in DRC and is a serious threat to sustained malaria control with pyrethroid LLINs. Next generation nets (PBO nets or bi-treated nets) should be considered for mass distribution.

Background
Malaria remains the leading cause of consultation, hospitalization, and death in the Democratic Republic of Congo (DRC), with on average more than 5,000 malaria deaths per month [1]. The National Malaria Control Program (NMCP) has a strategic goal of protecting at least 80% of the population at risk with preventative measures by 2020 [2]. The primary vector control method used to protect people in DRC is the free distribution of long-lasting insecticidal nets (LLINs). LLINs have been distributed on a provincial level, with rolling mass distributions scheduled for provinces approximately every three years, and routine distribution being done through ante-natal clinic (ANC) visits, expanded program of immunization (EPI) visits, and in some provinces, school-based distribution.  [16,17]. These results corresponded with the higher frequency of permethrin resistance than resistance to deltamethrin noted in susceptibility tests conducted in 2013 [18] (MPSMRM, 2014). Molecular analysis of pyrethroid resistant Anopheles malaria vectors in several locations in DRC has shown the upregulation of genes related to metabolic resistance that were associated with high rates of Plasmodium infection and loss of LLIN efficacy [19,20].
Previous published susceptibility data from DRC has focused on the use of a diagnostic concentration of insecticide to determine whether a mosquito population is susceptible or resistant [9, 18,21]. However, it is thought that the intensity of pyrethroid resistance may be an important indicator of potential pyrethroid LLIN control failure [22,23]. For this reason, annual insecticide resistance intensity testing has been scaled-up in DRC [18,21].
While agricultural use of pyrethroids has been associated with initial development of resistance in some studies [24,25], other studies have found that mass distribution of LLINs was associated with increasing resistance [26][27][28]. However, these studies have mostly looked at mosquito populations retrospectively and little is known about how rapidly these changes occur following a LLIN distribution campaign. Therefore, part of this study was to monitor pyrethroid resistance intensity in suburbs of Kinshasa before and after mass LLIN distribution in December 2016. Additionally, intensity of pyrethroid resistance was monitored nationwide in six other sites in 2017 and 11 sites in 2018.

Study sites
The first part of the study was conductedin Kinshasa Province in 2016 and 2017. Four sites were selected for mosquito larval collection to monitor changes in An. gambiae s.l. The second part of the study involved nationwide testing of pyrethroid resistance intensity. Deltamethrin and permethrin resistance intensity tests were conducted in seven sites in 2017 using Centers for Disease Control and Prevention (CDC) bottle bioassays. In 2018, testing was expanded to eleven sites, with resistance intensity to permethrin, alpha-cypermethrin and deltamethrin monitored using WHO tube tests for intensity For each round of bioassays, An. gambiaes.l. larvae were collected from the five sites ( Fig. 1) and returned to the laboratory at the Institut National de Recherche Biomédicale (INRB) in Kinshasa city, where they were reared to adults. Adult mosquitoes were kept in cages and provided with 10% sugar solution ad libitum until the time of testing at the age of 2-5 days.
The intensity assays conducted nationwide followed the same protocol, but tests were conducted once per year (all tests between January and August in 2017 and 2018) and mosquitoes were reared and tested in field insectaries.

CDC bottle bioassays
CDC bottle bioassays were conducted to determine the intensity of pyrethroid resistance, following standard guidelines [29,30]. Pre-measured vials of technical grade active ingredient were supplied by CDC and made into stock solutions for each insecticide dose by diluting with acetone. Stock solutions were stored in the refrigerator (4 °C) in lightproof bottles for future use. Glass Wheaton bottles (250 ml) were washed with warm soapy water and rinsed thoroughly with water at least three times and left to dry overnight. A disposable pipette was used to transfer 1 ml of acetone into the negative control bottle and 1 ml of each stock solution into the respective treatment bottle. Bottles were swirled so that the glass bottom and inside cap were coated before being placed on their side and rotated while rocking so that the sides were evenly coated with insecticide. The bottles were protected from sunlight, and caps removed before being left to dry overnight.
An aspirator was used to gently add twenty-five mosquitoes into each bottle per replicate.

WHO susceptibility tests
In 2018 insecticide susceptibility and resistance intensity testing was conducted in 11 sentinel sites (Fig. 1) using the WHO tube test. The insecticides tested in 2018 were: deltamethrin × 1, ×5, × 10 (0.05%, 0.25%, 0.5%); permethrin × 1, ×5, × 10 (0.75%, 3.75%, 7.5%) and alpha-cypermethrin × 1, ×5, × 10 (0.05%, 0.25%, 0.5%). In all sites, susceptibility testing was conducted with adult An. gambiae s.l., following WHO protocols [22]. INRB entomologists traveled to each site to collect larvae and pupae, which were reared to female adult mosquitoes aged 2-5 days and exposed for one hour to insecticidetreated filter papers provided by WHO (prepared by Universiti Sains Malaysia). All tests were accompanied by negative control tests where mosquitoes were exposed to filter papers impregnated with oil or solvent. Testing was done according to WHO protocols, with mortality read 24 hours after exposure. Four replicates of 25 An. gambiae s.l. were exposed to each concentration. determine the frequency of the voltage-gated sodium channel mutation (VGSC) 1014S (formerly known as kdr-east) and VGSC1014F (formerly known as kdr-west).
Genomic DNA was extracted from whole mosquitoes at CDC using Extracta™ DNA Prep for PCR-Tissue kits (QuantaBio, USA) and at INRB using the CTAB method. Species identification was performed according to the protocol of Wilkins et al. [31] at CDC and using the protocol of Santolamazza et al. [32] at INRB. The VGSC-1014S and 1014F alleles were detected using adapted protocols for allele-specific PCR (AS-PCR) [33][34][35]

Results
Intensity of resistance in Kinshasa Province before and after LLIN mass distribution using CDC bottle bioassay Over the four periods of testing, a total of 15,200 An. gambiae s.l. were used for resistance intensity bioassays in Kinshasa Province. Resistance to permethrin and deltamethrin was found in all sites (Fig. 3). In June 2016, before LLIN distribution mean results for Kinshasa (mean of Kingasani, Kinkole, Kimpoko, Bu) showed that permethrin resistance intensity was high and the mean mortality rate was 43% at the 5 × concentration and 73% at the 10 × concentration. After the mass distribution of LLINs in December 2016 (mean results for February, June and October tests) the mean mortality rate in Kinshasa was 32% for 5 × and 60% with permethrin at the 10 × concentration. The mean resistance intensity to deltamethrin was also high before LLIN distribution (75% at 5 × and 94% mortality at 10 × concentration) but decreased after LLIN distribution to a mean of 95% and 99% mortality at 5 and 10X concentrations, respectively. In general, levels of resistance were lower for deltamethrin, compared to permethrin. However, there was considerable variation in the results by site (Table 1). The hypothesis was that resistance intensity to permethrin and deltamethrin would increase in Kinshasa following LLIN distribution, compared to the control site of Kasangulu. In the control site of Kasangulu, mortality in permethrin intensity tests decreased significantly in 2017 (indicating an increase in resistance). Resistance intensity was greater in Kinkole for permethrin (OR 11.49, p-value < 0.001) and deltamethrin (OR 22.00, p-value < 0.001) compared to Kasangulu post-LLIN distribution and also in Kimpoko for deltamethrin (OR 2.57, p-value < 0.001). However, in Kingasani the opposite trend was recorded with a significantly lower resistance intensity following LLIN distribution for permethrin (OR 0.11, p-value < 0.001) and deltamethrin (OR 0.21, p-value < 0.001) compared to the control site, while in Bu there was no significant change post-distribution for both insecticides. Intensity of permethrin and deltamethrin resistance in six sites in DRC in 2017 using CDC bottle bioassay Nationwide bioassay testing showed that permethrin resistance was present in all 6 sites, with less than 10% mortality at the diagnostic dose. Mortality rates increased slightly with increased concentration, but high intensity permethrin resistance was present in all sites, with considerably less than 98% mortality at 10x the diagnostic concentration of permethrin (Fig. 4).
An. gambiae s.l. populations were less intensely resistant to deltamethrin, although all populations tested were resistant at 1x and 2x the diagnostic dose. The resistance intensity was low (> 98% mortality at 5x dose) in Kabondo and Inongo, moderate in Kalemie and Katana (> 98% mortality at 10x dose) and high (< 98% mortality at 10x dose) in Mikalayi and Kapolowe (Fig. 5).
One percent of samples did not amplify.

Discussion
Insecticide treated nets are believed to be an important source of selection pressure for pyrethroid resistance genes in African malaria vectors [27,36]. In addition to LLINs, other environmental factors such as agricultural pesticide run off into mosquito larval sites, exert additional selection pressure on malaria vectors [24,37]. Between 2011 and 2018, an estimated 134.8 million LLINs were distributed nationwide in DRC through mass campaigns and through routine distribution in schools and during ANC and EPI visits [38]. and Kinshasa in 2016 [18]. Resistance to permethrin, deltamethrin and alphacypermethrin now appears to be present nationwide.
An. gambiae s.s. was the most common vector species identified among the An. gambiae complex analyzed across 11 sites in DRC. However, in Kinshasa, An. gambiae s.s. (91%) and An.coluzzii (8%) were sympatric and there was a small proportion of hybrid An.
gambiae/coluzzii (5%) in Kinkole. Though the frequency of hybrids in the Anopheles population from Kinkole is low, mating seems to be occurring between the two species.
Both species are commonly sympatric in Central Africa, but hybrids of An. gambiae s.s/ An. coluzzii are usually very uncommon [42,43]. Populations of An. gambiae and An. coluzzii have previously been shown to be sympatric in several geographical areas in DRC, including Lodja, Mikalayi, Kalemie, Katana, Kinshasa, Kimpese and Inongo [18].
Both An. gambiae s.s. and An. coluzzii carried VGSC-1014F and 1014S alleles. However, An. coluzzii in Kinshasa had a high frequency of the 1014S allele, while An. gambiae s.s. had a high frequency of the 1014F allele; the latter observation may partially explain the higher levels of local permethrin resistance, despite deltamethrin-treated LLINs predominating in the most recent mass distribution campaigns. L1014F is suspected to play a larger contributing role in resistance to type I (permethrin) versus type II (deltamethrin and alpha-cypermethrin) pyrethroids [44]. Interestingly, a proportion of heterozygous An. gambiae s.s. from Kinshasa harbored both L1014F and L1014S alleles.
The phenomenon of these mutations co-occurring in individual mosquitoes has previously been reported in Senegal [45] and Uganda [46] and in Nord Ubangi, DRC [47]; however, the biological implications of possessing both resistance genotypes remain unknown and warrant further investigation.A limitation of this study is that only target site mutations for resistance were investigated. Mixed function oxidases (MFO) are implicated in pyrethroid resistance in several sites in DRC [48]. In addition, bioassays within 2016 showed increased mortality in permethrin resistant populations in DRC after pre-exposure to PBO [18].The genetic basis conferring resistance to pyrethroids in the DRC malaria vectors An. gambiae s.s. and An. coluzzii needs to be investigated at the national level to improve malaria control decision making, particular with regard to choice of LLINs for mass distribution campaigns.
Widespreadpyrethroid resistance,particularly high intensity resistance,is of great importance for the national malaria control program for the implementation of evidencebased resistance management strategies and deployment ofefficacious malaria vector control tools. Resistance intensity assays showed that neither 1, 5 or 10 times the diagnostic concentrations of permethrin, deltamethrin and alpha-cypermethrin were sufficient toprovide adequate mortality of An. gambiae s.l. collected from 7 nationwide sites in2017 tested using CDC bottle bioassaysand 11sites in2018using WHO tube tests.
WHO states that "when resistance is confirmed at the 5 × and especially at the 10 × concentrations,operational failure is likely" [22]. Pyrethroid LLINs should continue to offer some protection from malaria even in locations with high intensity resistance, through a combination of physical barrier, reduced survival of malaria vectors and malaria parasites [49][50][51]. However, next generation LLINs either impregnated with pyrethroids and the synergist PBO or containing chlorfenapyr (Interceptor G2® ) are potential alternatives for the improved efficacy of LLINs and for resistance management. Several experimental hut studies have shown improved efficacy of PBO and chlorfenapyr LLINsin controlling pyrethroid resistant malaria vectors compared to conventional pyrethroid LLINs [52][53][54][55][56].
LLINs containing PBO or novel insecticide classes should be considered by the National Malaria Control Program of DR Congo for future LLIN distribution campaigns in areas of moderate to high intensity of pyrethroid resistance, although the costs of these nets would also need to be considered.

Conclusion
The widespread presence ofmoderate to high intensity pyrethroid resistance across all sentinel sites in DRC is a great concern. Next generation LLINs with synergists or newinsecticide classesshould be consideredin DRC to sustain effective malaria control.

Funding
This study has been financially supported by U.S. President's Malaria Initiative

Availability of data and materials
All data generated or analysed during this study are included in this article and are available from the corresponding author.

Ethical consideration and consent to participate
The work described in this manuscript was determined to be non-human subjects' research by the CDC Center for Global Health (CDC 2016-242, 2016-082

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.