Patent Application
20140116112
The present disclosure relates to methods for rapidly determining the presence or absence of multiple contaminants in a test sample, such as a raw material intended for use in the preparation of a nutraceutical, using gas chromatography-mass spectrometry techniques.
As the use of nutraceuticals, such as multivitamins and other dietary supplements, has become more commonplace, concerns over the levels of purity, quality, consistency and potency of such supplements have increased. Ensuring that nutritional and dietary supplements are free of contaminants is particularly important when the supplements are intended for use by children and/or individuals with health problems, environmental sensitivities, etc. The US Food and Drug Administration (FDA) regulates dietary supplements as a category of foods and not as drugs, meaning that dietary supplements do not need to be specifically pre-approved by the FDA. In 2007, the FDA implemented a current Good Manufacturing Practices (cGMP) policy in an attempt to ensure that dietary supplements do not contain contaminants or impurities and are accurately labeled. However, the level of non-compliance with the cGMP is very high. Based on audits completed by the FDAs compliance division in 2011 and 2012, it has been estimated that nearly 70% of dietary supplement manufacturers are non-compliant with the cGMP policy. Significant concerns thus remain over the quality of nutritional and dietary supplements on the market today.
Many supplement companies obtain the raw materials for their supplements from a variety of suppliers, and then use those materials to formulate supplements for sale. Over 80% of the raw materials used in nutraceuticals sold in the US come from China and other non-US countries, leading to additional concerns over potential levels of contamination.
Methods that are typically employed to check for contaminants in raw materials, such as those used in nutraceuticals, are time consuming and expensive. There thus remains a need for methods that can rapidly and cost-effectively identify the presence or absence of, and/or determine the levels of, a large number of contaminants in raw materials intended for use in one or more dietary supplements.
The present invention provides methods for rapidly and accurately determining the presence or absence of, and/or quantifying the amount of, a large number of contaminants, such as pesticides, in a sample. In certain embodiments, the methods disclosed herein are employed to test for the presence or absence of contaminants, such as pesticides, in raw materials intended for use in nutraceuticals, such as vitamins and dietary supplements. Such raw materials include, but are not limited to, minerals and plant-based materials such as those listed in Table 1, below.
In one embodiment, methods for detecting the presence or absence of a plurality of contaminants in a sample are provided, such methods comprising: (a) extracting the sample with a water-miscible solvent in the presence of a high concentration of salts to provide a sample extract; (b) shaking and centrifuging the sample extract to provide a supernatant; (c) exchanging the water-miscible solvent in the supernatant for an organic, preferably non-water miscible, solvent methylene chloride to provide a treated supernatant; (d) analyzing the treated supernatant using gas chromatography-mass spectrometry (GC-MS) to provide a total ion chromatogram; (e) deconvoluting the total ion chromatogram to provide non-overlapping spectra; and (f) comparing the non-overlapping spectra with standard mass spectra for the plurality of contaminants, wherein the standard mass spectra are contained in a retention time-locked database. In certain embodiments, the water-miscible solvent is selected from the group consisting of: acetonitrile, ethyl acetate or acetone. In a preferred embodiment, the water-miscible solvent is acetonitrile. In certain embodiments, the organic non-water-miscible solvent is selected from the group consisting of: methylene chloride, hexane and toluene. In a preferred embodiment, the non-water-miscible solvent is methylene chloride.
Such methods can be used to quickly detect the presence or absence of at least 50, 100, 150, 200, 250, 300, 350, 400, 450, 500, 550, 600, 650, 700, 750, 800, 850 or 900 contaminants. In certain embodiments, the disclosed methods are employed to detect the presence or absence of a plurality of compounds selected from those listed in Table 2, below.
In one embodiment, an initial analysis is performed to determine whether or not one or more specific contaminants are present in a sample (i.e. to give a simple “yes or no” result). If this initial analysis indicates that the contaminant is indeed present, a second analysis is performed to determine the amount of the contaminant in the sample. In certain embodiments, the second analysis is performed using GC-MS.
As used herein, the term “deconvolution” refers to a mathematical technique that separates overlapping mass spectra (i.e. overlapping peaks in a total ion chromatogram (TIC)) into clean spectra of individual components.
As used herein, the term “high concentration of salts” refers to an amount of salts sufficient to provide a solution having a percentage composition by mass of salts between 40% and 90%, such as between 50% and 80% or between 60% and 70%. In certain embodiments, the term “high concentration of salts” refers to an amount of salts sufficient to provide a salt solution having approximately 65% composition by mass of salts. In one specific embodiment, the methods disclosed herein employ 4 g MgSO4, 1 g NaCl, 1 g trisodium citrate dehydrate and 0.5 g disodium hydrogen citrate sesquihydrate in 10 ml of solution.
As used herein, the term “nutraceutical” refers to food, or parts of food, that provide medical or health benefits, including the prevention and treatment of disease, and that are intended for consumption by a human or other mammal. The term nutraceutical encompasses, but is not limited to, dietary supplements including botanicals, vitamins, minerals, co-enzyme Q, carnitine, ginseng, gingko biloba, Saint John's Wort, saw palmetto, prebiotics and probiotics.
As used herein, the term “retention time-locking” refers to the matching of a first set of retention times obtained using a known chromatographic method having a defined set of column parameters and operating parameters to a second set of retention times obtained using a new, different, chromatographic method having a new, different, set of column parameters, wherein the second set of retention times are matched, or locked, to the first set of retention times.
As outlined above, the present disclosure provides rapid and cost-effective methods for detecting the presence or absence of multiple contaminants in a sample, such as raw materials for use in the preparation of a nutraceutical. Prior to analysis using GC-MS, the sample is extracted using a modified QuEChERS (Quick Easy Cheap Effective Rugged Safe) technique described in detail below. QuEChERS is a method for testing for pesticides that was developed by Michelangelo Anastassiades (Anastassiades et al., J. AOAC Int., 86:412-431 (2003)). This method entails solvent extraction of samples with acetonitrile, ethyl acetate or acetone, and partitioning with magnesium sulfate, either alone or in combination with other salts followed by clean-up using dispersive solid phase extraction (DSPE). More specifically, the sample is first extracted with a water-miscible solvent, such as acetonitrile, in the presence of a high concentration of salts (e.g. sodium chloride and magnesium sulfate) and buffering agents (e.g. citrate) to induce liquid separation and stabilize acidic and basic labile pesticides, respectively. After shaking and centrifugation, an aliquot of the organic phase is subjected to further clean up using DSPE. The resulting mixture is centrifuged and the resulting supernatant can either be analyzed directly or subjected to a concentration and solvent exchange step, if necessary, prior to analysis.
The extracted samples are then subjected to GC-MS analysis using methods well known to those of skill in the art and described below. The total ion chromatogram is deconvoluted as necessary using publicly available software, such as AMDIS (Automated Mass Spectral Deconvolution and Identification System; available from the National Institute of Standards and Technology (NIST)), Chemstation™ and/or DRS (Agilent Technologies, Inc. Santa Clara, Calif.).
The resulting spectra are compared with standard mass spectra for the contaminants of interest that are contained within a database that includes internal calibrations, such as a retention time-locked (RTL) database. Methods for automated retention time-locking are known in the art and include, for example those taught in U.S. Pat. No. 5,987,959. In certain embodiments, the resulting spectra are compared with those contained in a RTL pesticide database, such as the RTL Pesticide Library available from Agilent Technologies, Inc. This database contains locked retention time, compound name, CAS number, molecular weight and mass spectrum for 927 compounds, including pesticides, metabolite and endocrine disrupters, and other known contaminants. Using a RTL database eliminates the need to re-calibrate the GC-MS system for each potential contaminant and thus significantly reduces the time required to test a sample for the presence or absence of multiple contaminants.
Raw materials that can be analyzed using the methods disclosed herein include, but are not limited to, those shown in Table 1, below.
Table 2 shows a list of potential contaminants that can be detected using the methods disclosed herein, as published in Wylie, “Screening for 926 Pesticides and Endocrine Disruptors by GC/MS with Deconvolution Reporting Software and a New Pesticide Library” Application Note, Agilent Technologies, Inc., 2006.
The following examples are intended to illustrate, but not limit, this disclosure.
Approximately 1.0 g of sample was placed in a 50 mL tube and the exact weight was recorded on a log sheet. For each sample, two quality control samples were prepared using 1 g muffled sand; these were labeled “MB” (Method Blank) and “LCS” (Laboratory Control Sample). 9.0 mL of deionized water was added to each of the tubes. Quality control standards were added as follows:
(a) 50 uL of 20 ppm GC surrogate (tetrachlorometaxylene (TCMX), Decachlorobiphenyl (DCB), Tributyl phosphate and Triphenyl phosphate) in acetonitrile was added to all samples including the MB and LCS;
(b) 100 uL of 20 ppm OC pest spiking solution (Organochlorine Pesticide Mix AB #1 (Restek Corp., Bellefonte, Pa.) containing aldrin, α-BHC, β-BHC, δ-BHC, γ-BHC (lindane), cis-chlordane, trans-chlordane, 4,4′-DDD, 4,4′-DDE, 4,4′-DDT dieldrin, endosulfan I, endosulfan II, endosulfan sulfate, endrin, endrin aldehyde, endrin ketone, heptachlor, heptachlor epoxide (isomer B) and methoxychlor) in acetonitrile was added to the LCS; and
(c) 100 uL of 20 ppm Internal Standard solution in acetonitrile was added to all samples.
The samples were shaken vigorously and allowed to equilibrate for 2 hours at room temperature. Extraction of the samples was then performed by adding 10 mL of acetonitrile and shaking for one minute, adding the contents of an extraction salt packet (Q-Sep™ Q110 QuEChERS extraction salt packet containing 4 g MgSO4, 1 g NaCl, 1 g trisodium citrate dehydrate, 0.5 g disodium hydrogen citrate sesquihydrate; Restek Corp.), shaking again for one minute and then centrifuging for 5 minutes.
For samples needing clean-up (as determined for example, by previous difficulties with analysis, difficult matrix or darkly colored residues), the solvent extract was placed in a cleanup tube (Q-Sep™ dSPE 15 mL sample cleanup centrifuge tubes containing 900 mg MgSO4, 150 mg PSA and 45 mg GCB), shaken vigorously and centrifuged for 5 minutes, before being placed in an evaporation tube. The samples were then evaporated to near dry (less than 1 mL solvent) using a TurboVap™ evaporator, and 5 mL methylene chloride was added using a solvent pump. This process was repeated until the acetonitrile portion had been exchanged out for methylene chloride and the volume had reached less than 1 mL. Methylene chloride was then added to raise the volume in the sample back to 1 mL, and the sample was transferred into a labeled vial and cap using a crimper and aluminum cap.
Blanks were run with every set of samples to ensure that laboratory media or equipment was not leading to false positives in any contaminants. These were made following the same process as for the extracts for analysis.
Samples prepared as described above were analyzed for the presence of contaminants using an Agilent Technologies 5975C gas chromatograph/mass spectrometer (GC/MS) in combination with enhanced data analysis as described below.
Prior to analysis of samples, the GC/MS was checked for any instrument problems that could seriously affect the quality of analysis using routine procedures well known to those of skill in the art. Each analysis sequence carried out on the GC/MS was bracketed by calibration verification samples, “initial calibration verification” samples or “continuing calibration verification” samples. These samples were made using concentrations equal to 0.1 ppm. The concentration of the standards in these samples was within 50% of the expected values. The method blank (MB) and laboratory control (LCS) samples were placed at the beginning of the sequence.
The instrument was calibrated for target analytes, or contaminants, prior to reporting any target analyte concentration. Calibration was performed by running a set of samples containing a blank and five known concentrations, with the highest level corresponding to the highest expected results, through the screening method. The calibration set was quantitated using data analysis and deconvolution employing Deconvolution Reporting Software (DRS; Agilent Technologies, Inc). After each of the five calibration samples were quantitated, the new values were entered into the database. The curve shapes were checked for linearity. R̂2 values were 0.98 or greater.
Analysis was performed on completed data sets using DRS and Enhanced Data Analysis software (Agilent Technologies, Inc.). After files were deconvoluted, they were reviewed using QEdit™ software. Peaks identified by AMDIS (Automated Mass Spectral Deconvolution and Identification System; available from the National Institute of Standards and Technology (NIST)) and Chemstation™ software were reviewed for quality. More specifically, peaks were reviewed based on comparison between library spectra, AMDIS extracted spectra and Chemstation™ spectra; qualifier peak match; and retention time. Generally peaks were considered true “hits” if they had a MF (molecular formula) match value above 75 and the three qualifier ion values were within 25% of expected. In general, a genuine match has a spectrum very similar to the library/database spectrum, shows a strong, sharp peak shape in both the Chemstation™ and AMDIS peak viewer windows, has a very high MF value, and will likely be identified by both the Chemstation™ and AMDIS softwares.
There are times when Chemstation™ integrated a different peak than AMDIS. Because AMDIS has been rigorously developed to remove erroneous background suppression and use several statistical models to effectively “mine” the data, AMDIS was given higher priority than Chemstation™ when interpreting data. When a peak was accepted as genuine, it was either left alone or manually integrated. Since the GC/MS employed for these studies only uses a single quadrupole, it often has trouble resolving overlapping peaks. In such instances, the range of the correct peak was manually integrated. Once all of the peaks had been reviewed, the data was saved and a report generated.
Reports were reviewed after analysis to ensure that data met specifications, specifically sequence data information, internal standard recovery, surrogate recovery percentages, calibrated analyte concentrations and semiquant compounds (i.e. those compounds found using AMDIS and DRS that have not been calibrated for) hits were checked. Due to limitations with the software, generated reports often listed detections for compounds that had qualifier ion mismatches. These peaks do not show up in the default view and are very difficult to track down and delete, but are always erroneous. These were simply deleted from the report.
Internal standard recovery should have a minimum abundance of 1,000,000 counts, and should generally be within 50% of the calibrated abundance. Matrix effects can cause this number to vary somewhat, so data was generally accepted even if the recovery was outside of the 50% margin.
Surrogate recoveries should also be within 50% of calibrated values, but variation may also be due to matrix effects, thus this was not generally used to reject data unless there were clear signs that the ability to generate quality data was compromised.
Semiquant compounds that were found regularly in blanks were discarded from the screening list. These compounds, which included phthalates among a few others, were ignored when reporting data. Semiquant hits were not calibrated; unless they were calibrated, they can only be reported on a presence/absence basis.
In order for calibrated compounds to be reported, they must fall within the range of the initial calibration curve. If they were outside of this range, the sample was diluted to be within this range and rerun. Calibrated compound concentrations were multiplied by the dilution factor and divided by the sample weight before being reported.
The internal standard retention time was calibrated at 13.726 in accordance with the original AMDIS calibration. During the initial phase of calibration, the retention time was locked, which allowed AMDIS to accept or reject peaks based upon retention time. If the internal standard fell outside of the window and was not integrated by AMDIS, corrective action was taken and the sample was reanalyzed.
Sample cleanup was performed on samples that were excessively thick or colored, or that, based on previous history, were expected to cause problems during analysis.
All standards were stored at −4° C. or below and were allowed to reach room temperature before use.
The limits of detection (LOD) determine the lowest concentration at which an analyte can be detected in an extracted sample. Since these measurements are not available for all compounds, the average of the LODs for calibrated compounds determines the estimated detection limit for uncalibrated compounds. The LOD is determined by the lowest concentration compound extracted with a signal to noise ratio of 2.5-5. These tests were performed periodically to determine any changes in instrument sensitivity.
Initial calibration established a calibration curve used to determine the concentration of calibrated compounds and recoveries of surrogates. The average internal standard response was also used to determine the baseline response used for the internal standard calibration verification. Calibrations were run at seven levels: 0.01, 0.025, 0.05, 0.10, 0.5, 1.0, and 5.0 ppm. Using the data from this calibration, each compound should have a linear or quadratic curve with an R sq. value of 0.95 or greater. The lowest calibration level determines the limit of quantification (LOQ). If a compound failed calibration (i.e. did not have an R sq. value of 0.95 or greater) it was noted and corrected before any detections of this compound were quantitated.
Internal standard calibration verification (ISCV) was used to verify instrument performance and internal standard response. This was prepared with 1.0 ppm internal standard in methylene chloride. The internal standard abundance should be 70%-170% of the response established with the initial calibration and within the AMDIS retention time window. If the response fell outside of this window, the aberration was investigated and corrected before analysis took place. The ISCV was also used to determine column condition. The abundance of ion 207 (siloxane bleed) at 40 minutes should be under 25,000. If the background did not improve in subsequent analyses, the column was replaced and the instrument recalibrated before sample analysis.
Initial calibration verification (ICV) and continuing calibration verification (CCV) samples were used to verify that the analysis performance was within the parameters of the initial calibration. The CCV was run at the end of a set of samples to bracket either an initial calibration or ICV sample. The ICV was run in place of a set of calibration samples unless there were measures outside of limits requiring a new calibration set.
Recovery control limits for surrogates and other calibrated compounds were set at 70-170%. These laboratory control spikes calculated percent recovery. If these fall outside of limits, it could be due to matrix suppression, problems with analysis or extraction. These issues were addressed as necessary.
Sodium ascorbate intended for use in nutraceuticals for human consumption was tested for the presence of multiple pesticide residues as described above. No pesticide residues in amounts above the USP <561> Articles of Botanical Origin reporting limits were found.
While the present invention has been described with reference to the specific embodiments thereof, it should be understood by those skilled in the art that various changes may be made and equivalents may be substituted without departing from the true spirit and scope of the invention. In addition, many modifications may be made to adapt a particular situation, material, composition of matter, method, method step or steps, for use in practicing the present invention. All such modifications are intended to be within the scope of the claims appended hereto.
All of the publications, patent applications and patents cited in this application are herein incorporated by reference in their entirety to the same extent as if each individual publication, patent application or patent was specifically and individually indicated to be incorporated by reference in its entirety.
This application claims priority to U.S. provisional patent application No. 61/718,607, filed Oct. 25, 2012.
| Number | Date | Country | |
|---|---|---|---|
| 61718607 | Oct 2012 | US |
