A New Method for Ivermectin Detection and Quantification through HPLC in Organic Matter (Feed, Soil, and Water)

Ivermectin is a macrocyclic lactone widely used in veterinary medicine for its broad-spectrum antiparasitic properties. It has been proven to be effective and safe. The purpose of this study was to develop a high-performance liquid chromatography method with a diode array detector for ivermectin screening in feed and water for animal consumption. Furthermore, the objective was to quantify ivermectin levels that were higher than 0.5 mg/kg in solid matrixes and 0.1 mg/kg in water. Doramectin was used as process standard. Samples were extracted using solid phase extraction with silica and C-18 columns. The method involved the use of high-performance liquid chromatography (HPLC) with a diode array detector (DAD). The results were interpreted using a calibration curve built with ivermectin standards at multiple concentrations (0.5, 1, 2, 5, and 12.5 mg/kg). Statistical evaluation of data was done using ANOVA. The data analysis showed that the linear regression was highly significant (P < 0.001), the intercept values were not significantly different from zero, and the correlation coefficient values (>0.999) indicated excellent linearity. Further tests demonstrated that this method is also useful when studying soil matrixes. The soil was dried and analyzed in the same way as feed; the same recoveries were realized on the spiked samples. The method is easy, inexpensive, precise, and repeatable; it requires very small amounts of sample.


Introduction
Ivermectin is a macrocyclic lactone widely used in veterinary medicine for its broad-spectrum antiparasitic properties. Ivermectin is naturally produced by the yeast Streptomyces avermitilis, and it consists of a mixture of two compounds: 22,23-dihydroavermectin β1a (C 48 H 74 O 14 , B 1a ) and 22,23dihydroavermectin β1b (C 47 H 72 O 14 , B 1b ). Ivermectin B 1a makes at least 80% of the mixture, while no more than 20% is made of ivermectin B 1b . It is active against both arthropods and nematodes [1,2], causing their paralysis and subsequent death by blocking chloride channels [3]. Like all avermectins, ivermectin is highly liposoluble [4].
Ivermectin is authorized for oral, topical, and parenteral (subcutaneous) administration [1,2,5]. Te European Medicine Authority has authorized ivermectin for multiple species: cattle, pigs, sheep, goats, horses, reindeers, mice, rats, dogs, cats, primates, and humans. Ivermectin is marketed either alone or in combination with other antiparasitic compounds [6,7]. In farm animals, doses can vary according to the route of administration and range from 200 µg per kg of body weight in cattle, sheep, and goats to 300 µg per kg of body weight in pigs [6,8].
It is estimated that between 80 and 98% of the drug is excreted in feces, without being metabolized, thus reaching the environment intact.
Maximum residual limits (MRLs) have been established for all mammalian food-producing species, except animals producing milk for human consumption. Te marker residue is 22,23-dihydroavermectin B1a. Te MRLs are restricted in the targeted tissues in all species as follows: 30 µg/ kg in the kidney and 100 µg/kg in the fat and liver [8,9]. Te Italian National Reference Center for the surveillance and control of feed for animals (CReAA-Centro di Referenza Nazionale per la Sorveglianza e il Controllo degli Alimenti per gli Animali) has established a detection limit of less than 1 mg/kg in feed.
Although ivermectin has a wide safety margin in mammals [1,2], it has been shown to be toxic in numerous invertebrate species. Mesa et al. reported possible ecological consequences in aquatic ecosystems [10]. Johnson-Arbor has raised environmental concerns about ivermectin residues [11]. Terefore, an accurate system to monitor ivermectin levels in organic substrates would therefore be of crucial importance.
Diferent methods for ivermectin detection have been reported by several authors.Åsbakk et al. used highperformance liquid chromatography (HPLC) with fuorescence detection to detect ivermectin in feces from reindeers [12]. Iglesias et al. used the same method for soil and cattle feces [13]. Liquid chromatography-tandem mass spectrometry (LC-MS-MS) has also been shown in several studies to be useful for multiple matrixes such as plasma, milk, liver, and whole blood [14][15][16][17][18][19][20]. Liquid chromatography-electrospray ionization-mass spectrometry/mass spectrometry (LC-ESI-MS/MS) has also been used to detect macrocyclic lactones in fsh muscle [14].
Tis study aims to develop a HPLC method with a diode array detector (DAD) for screening for ivermectin in feed and water for animal consumption. Furthermore, we wanted to quantify ivermectin levels higher than 0.5 mg/kg in solid matrixes and 0.1 mg/kg in water. We used doramectin as process standard [13]. . Of the two scales used, one had 10 mg sensitivity and one had 0.01 mg sensitivity.
Glassware used consisted of 10 mL v-bottom graded test tubes, 10 mL u-bottom test tubes, 10 mL-20 mL amber test tubes with screw caps, class A or B glass pipettes, Pasteur glass pipettes, graded cylinders, class A fasks, and amber vials.

Chemicals and Reagents.
All reagents had known analytical purity. Te water used was deionized and ultrafltrated water.
Methanol for analysis from Carlo Erba Reagents (DASIT Group S.p.A., Milan, Italy) was used for the extraction, while Aluminium oxide 90 (neutral alumina) from Merck (Merck KGaA, Darmstadt, Germany) was used for the preliminary chromatography phase. Te following solvents were used for SPE column activation and elution as well as components of the HPLC mobile phase: methanol for HPLC from Carlo Erba Reagents (DASIT Group S.p.A., Milan, Italy), acetonitrile for HPLC from Carlo Erba Reagents (DASIT Group S.p.A., Milan, Italy), and dichloromethane for analysis from Carlo Erba Reagents (DASIT Group S.p.A., Milan, Italy). Ivermectin with declared purity from Dr Ehrenstorfer (LGC Labor GmbH Augsburg, Germany) and doramectin with declared purity from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany) were used, respectively, for the fortifed specimen and as an internal standard.
For each new experiment, a mixture of solvents for dilution made of acetonitrile (56%), methanol (37%), and water (7%) (v/v) and the activated neutral alumina (100 g of neutral alumina moisturized with 6 mL of deionized water) were prepared before use.
All solvents and solutions used for the HPLC system were fltered on a 0.2 μm flter.

Preparation of Standard and Quality Control Samples.
All standards and controls were prepared just before use. A stock solution of ivermectin and one of doramectin was prepared in a 10 mL fask, considering the declared purity, at a concentration of 1000 µg/mL. Te solution was made with methanol for HPLC and stored in the freezer in an amber tube for no more than 12 months [21].
An intermediate solution was made for each compound at 10 µg/mL, starting from the stock solution. In a 10 mL fask, the stock solution was added to the solvent mixture (acetonitrile (56%), methanol (37%), and water (7%)) at a concentration of 1 : 100.
From the intermediate solutions, ivermectin was diluted with the solvent mixture to obtain fve working solutions at 0.1, 0.2, 0.4, 1, and 2.5 µg/mL for calibration points. Another working solution (0.05 µg/mL) was made by adding 50 µL of the intermediate solution of ivermectin and doramectin to 9.95 mL of the solvent mixture.
For each analytical session, a blank control and a fortifed sample at 0.1 mg/L were set up for water samples. In a 10 mL fask partially flled with deionized water, 100 µL of intermediate solution at 10 µg/mL was added, made up to volume, and diluted in a ratio of 1 : 2 with a dilution mix.
For each analysis session, a blank control and a fortifed control were set up for solid samples (feed or soil). Negative control (blank) consisted of 10 g of an ivermectin-free sample (feed or soil). Te fortifed sample (0.5 mg/kg) was made of 10 g of an ivermectin-free sample (feed or soil) to which 500 µL of the intermediate solution at 10 µg/mL was added and left for at least 10 minutes.
All samples (controls and tests) were subjected to the same extraction, purifcation, and analysis procedure.

Experimental Design.
Te solid samples were extracted with methanol and purifed using a solid phase extraction (SPE) technique. Te purifed extracts were analyzed using HPLC after dilution.

Extraction and Purifcation of the Sample.
Te water samples were diluted in a ratio of 1 : 2 with the diluting solution, and 500 µL of the intermediate solution of the internal standard (doramectin) was added at 10 µg/mL; 1.5 mL was transferred in an amber vial.
Soil samples were dried in a heater at 45°C prior to processing. At least 2/3 of the total solid samples (feed or soil) were shredded to obtain a homogeneous powder. Using a scale with a sensitivity of 0.01 g, 10 g of ground samples were transferred to a conical fask. 50 mL of methanol and 500 µL of the intermediate solution of the internal standard (doramectin) at 10 µg/mL were added to the sample, and the fask was capped. Te fasks were placed in a water bath with ultrasound for 20 minutes and then shaken on the shaker for one hour at 700 oscillations/minute. 40 mL of supernatant was placed in plastic test tubes with a pressure cap and centrifuged at 3000 rpm for 5 minutes. A frit was put inside each 6 mL reservoir, then 5 g of activated neutral alumina were added, and another frit was put on top. Te 50 mL reservoirs were then fxed on top with the appropriate adapters. 15 mL of centrifugate supernatant was added to the 50 mL reservoir. Te frst 5 mL was discarded, and the remaining volume was collected in glass tubes.
For activation, the SPE silica columns were inserted into the manifold, one for each sample plus the blank and fortifed controls. We percolated 5 mL of acetonitrile for HPLC by gravity in columns and then dried them for 1 minute using vacuum. 5 mL of dichloromethane for HPLC was leached and dried for 1 minute after applying vacuum.
For conditioning, the SPE C-18 columns were inserted into a second manifold, one for each sample plus the blank and fortifed controls. 5 mL of dichloromethane was percolated by applying vacuum and dried for 1 minute; then 5 mL of a water/acetonitrile mixture (50 : 50 v/v) was percolated, and the column was kept wet.
A 10 mL syringe was applied to the SPE C-18 columns using the appropriate adapter. Simultaneously, 5 mL of the water/ acetonitrile mixture (50 : 50 v/v) and 2 mL of each sample were put in each syringe using a calibrated class A pipette. Te fuid was allowed to percolate, keeping the column wet. Columns were then washed with 1 mL of the water/acetonitrile mixture (50 : 50 v/v) that was left to percolate by gravity and then dried with vacuum for 15 seconds. Te walls of the columns were washed by adding 2 mL of the water/acetonitrile mixture (50 : 50 v/v) and by letting it fow with the use of the vacuum pump. Finally, the column was dried for 1 min, maintaining the vacuum.
Te SPE C-18 columns were inserted into the conditioned SPE silica columns with the appropriate adapter. 5 mL of the dichloromethane/acetonitrile mixture (9 :1 v/v) was added to the SPE C-18 columns and left to percolate, maintaining the vacuum for 1 minute; another 5 mL of the mixture was subjected to the same procedure. Te SPE C-18 columns were discarded, and ivermectin was eluted from the silica SPE column with 5 mL of acetonitrile for HPLC by gravity, without letting it go dry. A further 2 mL of acetonitrile for HPLC was added and eluted by applying vacuum. Te eluate was collected in a conical tube and dried in a nitrogen stream at 50°C ± 5°C. Te eluate was then diluted with 2 mL of the solvent mixture and vortexed for 1 minute, and the samples were fltered on 0.45 pm flters. Approximately 1.5 mL was placed in amber vials.

2.4.2.
Analysis by High-Performance Liquid Chromatography. Reverse-phase liquid chromatography was carried out. Te mobile phase consisted of a solution containing acetonitrile (56%), methanol (37%), and water (7%) in an isocratic process. Te stationary phase consisted of a C-18 precolumn and a C-18 column, respectively, a Supelguard reverse-phase chromatographic column C-18 (2 cm long, 4.6 mm internal diameter, and 5 µm average particle diameter) and a C-18 Supelcosil reverse-phase chromatographic column (25 cm long, 4.6 mm internal diameter, and 5 µm average particle diameter).
All HPLC analyses were conducted at room temperature (20°C). Before injecting the samples, the system was equilibrated for 30 minutes under the conditions described above. Te fow was set at 1.2 mL/min, and the wavelength at 245 nm. A volume of 100 µL was injected for each sample run.
Te ivermectin standard (0.05 µg/mL), the internal standard (0.05 µg/mL), the blank (Figure 1), and the fortifed controls ( Figure 2) were analyzed frst. Subsequently, all samples were run and the standard injections were repeated every three samples.
Te analysis was continuously monitored by Accredia, the Italian accreditation body, which certifes the quality of laboratory practices and conformity of the results obtained.

Results and Discussion
As can be seen in Figure 3, the calibration line had the nanograms injected on the x-axis and the area under the peak on the y-axis. Te calibration curve and instrument sensitivity are acceptable when (1) Te correlation of determination (r 2 ) is greater than or equal to 0.99. (2) Te peak relative to the 1 st point of the calibration curve is quantifable.
Te concentration of ivermectin present in the sample was calculated based on the photometric response obtained from the chromatogram report, as can be seen in Figure 4. In the case of positive samples above the maximum concentration, the sample was diluted to remain in the scope.
Te concentration of ivermectin in the sample was calculated as shown in equation (1), where A is the injected amount of ivermectin (ng), P is the weight of the sample (g), and 2 is the result obtained from the simplifcation of the calculation: ng injected x 1000 µI x 50 mL x 2 mL/100 µL x P x 2 mL x 1000. . (1) Te concentration of ivermectin added to the fortifed sample was calculated by interpolation on the calibration line.
To qualitatively confrm the presence of ivermectin, cochromatography was carried out to make sure that the analyte in the suspect sample is indeed the one that is being investigated. Te chromatography is run by injecting 50 µL of a known sample extract together with a 50 µL volume of one of the working solutions of ivermectin. Te working solution is chosen so that it corresponds to the same amount in ng of analyte comparable to that present in the sample. Te perfect overlap of the standard and sample confrms the association when the peak has a good shape, and it is symmetrical without splitting or any deformations that may indicate the presence of an interferent and not of an analyte. Te area under the peak will be equal to the sum of half of the standard and half of the sample because 50 µL of each was injected instead of a full injection of 100 µL. Te spectrum of the sample peak was also compared with the spectrum at intermediate concentration, evaluating the overlap factor (match factor), which must not be less than 900/1000.

Precision and Accuracy.
Te quantitative confrmatory method validation was performed according to the Annex III of the EU Regulation 2017/625. Te measurement uncertainty was expressed, at the concentration subject to validation (detection limit of 0.5 mg/kg), using the β error (false negative) result less than or equal to 5% (statistical certainty 1 − β � 95%).

Specifcity, Detection, and Quantifcation.
Specifcity was assessed by applying the entire test method to 20 blank samples for both solid and liquid samples. None of the samples showed chromatographic peaks at the analyte retention time.
To verify detection capability, extracts of 20 blank sample and the relative 20 strengthened at a value of 0.5 mg/kg were treated and injected in 2 analytical sessions. In all the fortifed samples, the chromatographic peak at the ivermectin retention time was highlighted, while the blank samples did not show any chromatographic peak at the ivermectin retention time. Terefore, the maximum tolerable error of 5% was respected. Te limit of detection (LOD) was set at 0.25 mg/kg and tested experimentally.  Journal of Analytical Methods in Chemistry Te limit of quantifcation (LOQ) was established at 0.5 mg/kg. It was experimentally verifed through tests on fortifed feed at the limit of quantifcation in six independent replicas.

Accuracy.
Accuracy was assessed through the 3-level recovery rate within the scope of the method. Precision was assessed by calculating the standard deviation and the coefcient of variation (%) by conducting 6 replicate tests for each of the concentration levels.
Linearity was evaluated to understand the feld of application of the method (0.5-10 mg/kg). During calibration, r 2 obtained was 0.9999 for ivermectin. Specifcity was assessed by applying the entire test method to 6 blank samples. Te tests conducted made it possible to verify the absence of an instrumental response around the retention time.

Uncertainty.
Uncertainty was expressed as the maximum extended relative uncertainty. In expressing the uncertainty, the actual degrees of freedom (v), calculated using the Welch-Satterthwaite formula, the calculated coverage factor (k), and a 95% confdence level (p) were considered. Tese parameters are summarized in Table 1.
From the above uncertainty, we calculate the uncertainty of the result (C) using the following equation:

Method Validation.
Te method was validated for both screening and confrmatory purposes. For the screening validation, the selectivity, specifcity, and quantifcation limit were verifed by the 20 independent analyses of blank feed samples and 20 independent analyses of blank feed samples spiked at the quantifcation limit concentration settled at 0.5 mg/kg. Te same was done for drinking water: 20 independent analyses of blank water samples and 20 independent analyses of blank water samples spiked at the quantifcation limit concentration settled at 0.5 mg/kg. Te method was found to be applicable also to this matrix.
Te applicability of the method was studied for feed matrixes, and the concentration range was settled from the limit of quantifcation, 0.5 mg/kg, up to 10 mg/kg, considering the content of ivermectin in medicated feed. Further tests demonstrated that this method is also useful when studying soil matrixes. Indeed, when the soil was dried and analyzed in the same way as feed, sufcient recovery rates were met in spiked sample.
Linearity was evaluated to cover the entire range of application of the method. For each calibration curve (Figure 3), fve concentration levels were chosen (0.5, 1, 2, 5, and 12.5 mg/kg) and every level was replicated three times. Te calibration curves were obtained using least squares Journal of Analytical Methods in Chemistry regression analysis. Te dataset obtained was investigated using the ANOVA approach. Te elaboration showed that the linear regression was highly signifcant (P < 0.001), the intercept values were not signifcantly diferent from zero, and the correlation coefcient values (>0.999) indicated excellent linearity. Selectivity and specifcity were evaluated by analyzing six samples from diferent kinds of feed and by confrming the absence of analytical substances with retention times similar to ivermectin.
Te detection limit was fxed and verifed at half of the LOQ value by analyzing six spiked feed samples at 0.25 mg/kg. Te quantifcation limit was fxed and verifed at 0.5 mg/ kg at half of the minimum quantifcation level concentration recommended by CReAA. Other validated methods involving LC in diferent matrixes have reached lower sensitivity by using diferent detectors, for example, MS was used by Morbidelli et al. in plasma to reach a 0.5-20.0 ng/mL range [17], by Durden et al. who saw in dried blood spots concentration of 1 ng/mL [16], and by Duthal et al. who could detect a range of e 0.5-60 ppb in milk [20].Åsbakk et al. and Iglesias et al. used a fuorescence detector reaching sensitivity at 5−2000 ng/g and 1-2000 ng/g, respectively [12,13].
To determine trueness, precision, recovery, and repeatability, spiked blank feed samples were used because no certifed reference material was available for ivermectin. In detail: blank feed samples were spiked, before the beginning of the extraction procedure, with ivermectin at three different concentration levels, and 6 independent replicas were carried out for each spiking level.
Te results obtained are shown in Tables 2 and 3. Ruggedness and reproducibility were evaluated using ongoing data. Ruggedness was verifed by comparing processed quality control samples made by diferent operators under diferent environmental conditions and at diferent times with diferent batches of standards and laboratory materials. Te analytical method was found to be rough, as the quality controls results are stable under the planned operating conditions. Te reproducibility CV%, calculated from 40 ongoing quality control data, is 8.40%.

Conclusions
Te HPLC method presented here allows the detection and quantifcation of ivermectin levels in diferent matrixes by creating a calibration curve and using doramectin as a process standard. Tis validated method is easy, cheap to use, precise, and repeatable, and it needs very small amounts of samples.
It could be easily used to detect ivermectin in solid and liquid matrixes for monitoring and environmental purposes. In fact, this method could be a reference method for

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.