Quantification of carnitine levels in dialysis patients

Description of equivalent WO 2004026114 A2

QUANTIFICATION OF CARNITINE LEVELS IN DIALYSIS PATIENTS

CROSS REFERENCE TO RELATED APPLICATIONS

This application claims priority under 35 U.S.C. §119(e) to U.S. Patent Application Serial No. 10/252, 115 entitled "Quantification of Carnitine Levels in Dialysis Patients" filed September 23, 2002, which is a continuation-in- part of U.S. Application Serial No. 09/464, 132, filed December 16, 1999, now U.S.

Patent , ,, issued September 24, 2002, which application is a continuation-in-part of application Ser. No. 09/277,119, filed March 26, 1999, which in turn claims priority from provisional application Ser. No. 60/117,880, filed January 30, 1999, the disclosures of all applications being entirely herein incorporated by reference.

FIELD OF THE INVENTION

The present invention relates to methods for quantifying the level of carnitine deficiency in patients such that carnitine concentrations can be easily and accurately tracked within a given patient over time. In particular, the present invention pertains to methods for quantifying the carnitine concentrations and the level of carnitine deficiency in patients undergoing dialysis procedures.

BACKGROUND OF THE INVENTION Carnitine is a naturally occurring substance in the human body required for energy metabolism at the cellular level. It has been shown to have a role in transporting fatty acids into mitochondria to help produce energy and in removing toxic waste from the cells. Studies indicate that more than 70% of the carnitine present in the plasma of a hemodialysis patient is removed during a single dialysis session. This high percentage carnitine loss by a patient during a hemodialysis session is thought to be attributable to the compound's relatively small molecular weight, high water solubility, and poor protein binding. It is believed that carnitine levels are further diminished in end stage renal disease ("ESRD") patients by reduced renal synthesis and reduced intake of meat and dairy foods. Thus, the carnitine levels of dialysis patients, and end stage renal disease patients undergoing dialysis in particular, generally decrease steadily over time.

The depletion of carnitine levels in tissue is a noted long-term consequence of repeated losses of carnitine from plasma. Thus, carnitine depletion in tissue is experienced by patients that undergo successive dialysis sessions over a prolonged period of time. Tissue carnitine depletion is undesirable because it is associated as a possible causal factor for many health complications associated with dialysis, including cardiomyopathy, arrhythmias, muscle weakness, and general fatigue.

As described in U.S. Patents 6,335,369 and 6,429,230 to Cavazza (henceforth, "the Cavazza patents"), the administration of levocarnitine (or L- carnitine) may be beneficially used to treat carnitine deficiency in ESRD patients (also known as "chronic uremic patients") undergoing regular dialysis. According to the Cavazza patents, ESRD patients undergoing periodical dialysis can be administered levocarnitine or one of its salts to prevent or treat carnitine deficiency. This is taught to be accomplished by, among other things, administering an effective dose of L-carnitine intravenously via a venous return line after each dialysis session of the ESRD patient. The Cavazza patents teach that initiation of such L-carnitine injection therapy to treat carnitine deficiency may be prompted by patients demonstrating pre-dialysis plasma carnitine concentrations that are below normal (normal carnitine concentration levels are 40-50Dmol/L). Such intravenous administration of L-carnitine to ESRD patients undergoing dialysis treatments is taught to result in increased plasma carnitine concentrations. Additionally, carnitine deficiency in ESRD patients who are undergoing dialysis can be prevented by the intravenous administration of levocarnitine prompted by carnitine concentration levels in plasma that are trending toward carnitine deficiency. The methods for the prevention of treatment of carnitine deficiency as taught by the Cavazza patents thereby make it possible to correct for the loss of plasma carnitine which otherwise could take place during a dialysis session.

In order to manage the therapeutic delivery of L-carnitine to dialysis patients, it would be beneficial for physicians to have an accurate and reliable, yet fast and low cost, method to monitor the level of plasma carnitine in patients before a given dialysis session, after a given dialysis session, and/or during a given dialysis session. The current techniques available for measuring carnitine levels in a patient's plasma include enzymatic assays, high-pressure liquid chromatography ("HPLC"), and tandem mass spectrometry ("MS/MS"). These techniques for quantifying plasma carnitine levels, however, do not generally provide the high throughput, low cost, and high sensitivity and accuracy analysis necessary for implementation on a large-scale basis to manage levocarnitine replacement therapies for patients undergoing dialysis.

Each of the contemporary techniques for measuring carnitine levels in patients suffers from distinct disadvantages. Enzymatic assays are highly sensitive, but are extremely time consuming to run and, therefore, are not readily adaptable to large-scale use by dialysis clinics. Furthermore, enzymatic assays can only provide quantification of free carnitine (L-carnitine) and total carnitine levels in a plasma sample and do not provide results for individual acylcarnitine s, such as long chain carnitines, which are useful markers of the metabolic status of ESRD patients undergoing dialysis for selecting appropriate therapies. Furthermore, the determination of total carnitine levels using enzymatic assays must be carried out separately from the determination of free carnitine levels due to the necessary use of chemical hydrolysis prior to quantification. HPLC methods involving specific pre-column derivatization and fluorescence detection have been demonstrated to be sufficiently sensitive to allow the determination of levels for several individual acylcarnitine s. However, HPLC methods remain expensive and very time consuming, in large part because the sample preparation require elaborate, multi-step procedures. Furthermore, the determination of long chain and very long chain acylcarnitine levels cannot be carried out by HPLC. HPLC further suffers from the fact that the quantification of total carnitine levels requires a separate analytical run including chemical hydrolysis of the sample.

Notably, both the HPLC and enzymatic assays methods that can be used for accurate and reliable determination of carnitine levels require biological samples that have been kept frozen during storage and shipment following collection. Understandably, this factor makes remote laboratory utilization of these procedures difficult.

With regard to MS/MS quantification of carnitine levels, electrospray ionization MS/MS has been used to identify abnormal carnitine profiles in infants from blood samples. In particular, U.S. Patent No. 6,258,605 to Chace ("the '605 patent") discloses a method for genetic screening of infants using tandem mass spectrometry ("MS/MS"). According to the '605 patent, blood is taken from an infant patient and spotted in small amounts on a piece of filter paper where it is dried and sent to a lab. The blood is then reconstituted at the lab and run through an electrospray ionization mass spectrometer with an internal standard to identify potentially abnormal metabolic profiles of amino acids and acylcarnitine s. By collecting dried blood samples on filter paper, the method taught by the '605 patent allows for inexpensive, un-refrigerated transfer of blood samples from the doctor to a remote lab for analysis. This in turn permits a single, remotely located mass spectrometer to be used to handle a large volume of samples, thereby taking advantage of economies of scale.

Although electrospray ionization MS/MS of dried blood samples as taught by the '605 patent is a reliable, low cost and high throughput mechanism for identifying abnormal metabolic profiles of acylcarnitines in infants, it is not a suitable solution for obtaining accurate quantifications of plasma carnitine levels in carnitine deficient patients for several reasons. First, blood can be a poor indicator of tissue carnitine levels because the intracellular concentration of L- carnitine in red blood cells does not equilibrate with the plasma concentration (the membranes of the erythroeytes being impermeable to L-carnitine). Only the carnitine present in a patient's plasma can be dialyzed during a given dialysis session while the carnitine concentration within the red blood cells is not directly affected by dialysis. Furthermore, since red blood cells do not have mitochondria, carnitine cannot be utilized therein for the D-oxidation of long-chain fatty acids. Therefore, the intracellular concentration of carnitine present in red blood cells may in fact skew any analyses for carnitine deficiency. Preparation of dried blood samples for introduction into the mass spectrometer as taught in the '605 patent would necessarily release a large portion of carnitine from the red blood cells and thereby overstate the determination of free and total carnitine levels.

In addition, the derivation steps employed during sample preparation for many MS/MS analysis methods are known to cause hydrolysis of acylcarnitines into free carnitine, as has been reported by Johnson et al. (1999) among others. This hydrolysis of acylcarnitines can cause over-reporting of free carnitine. Unfortunately, such over-reporting of free carnitine is clinically significant if the method were to be used to manage carnitine replacement therapy for dialysis patients. Prior solutions to this problem entail running an MS/MS assay twice on a given sample, once without derivatization to measure free carnitine levels and a second time with derivatization to measure total carnitine levels. This, however, increases sample analysis costs and therefore loses a significant benefit of MS/MS methods. Furthermore, while MS/MS can be employed without a derivatization step, removal of this step causes the sensitivity of the assay to decrease over ten fold. This marked decrease in sensitivity is problematic in the quantification of carnitine deficiency because the screened samples taken from carnitine deficient patients undergoing dialysis are necessarily expected to have low levels of carnitine present.

A further complication to the adaptation of MS/MS to quantifying carnitine levels is that glutamic acid is believed to interfere with the analysis of acetylcarnitine concentrations. Tandem mass spectrometry is capable of distinguishing molecular ions of the same mass that also have different product ions. Tandem mass spectrometry, however, cannot distinguish between molecular masses having the same product ions and molecular ions. The di-butyl ester of the protonated from of glutamic acid shares the same molecular ion molecular mass (260 m/z) and product ion molecular mass (85 m/z) as the butyl ester of acetylcarnitine. Since glutamic acid in physiological concentrations is much more abundant than acetylcarnitine, the protonated form glutamic acid can be present in sufficient amounts that can interfere with the MS/MS signal at 260 m/z which is associated with the detection of acetylcarnitine. Therefore any utilization of MS/MS for analysis of carnitine concentrations must account for this problem. Therefore, there remains a need in the art for an improved method for diagnosing carnitine deficiency in patients and quantifying the level of carnitine deficiency such that carnitine concentrations can be easily and accurately tracked within a given patient over time. Also, there remains a need in the art for a high throughput, low cost, and high accuracy method for quantifying the level of carnitine deficiency in a patient such that the method can be used frequently to monitor the level of carnitine deficiency and manage appropriate therapies to treat carnitine deficiency.

SUMMARY OF THE INVENTION In light of the above-described deficiencies present in contemporary techniques for diagnosing and quantifying carnitine deficiency in patients undergoing dialysis, it is an object of the present invention to provide a reliable and accurate method for quantifying carnitine levels in dialysis patients.

Furthermore, it is an object of the present invention to provide a method for a high throughput, low cost, and high accuracy quantification of the level of carnitine deficiency in patients to help manage therapies for treating dialysis atients.

Additionally, it is an object of the present invention to provide a highly sensitive method for determining free carnitine, acylcarnitine and total carnitine levels in dialysis patients. Such levels can be utilized to track patient health and assist in diagnosing carnitine deficiency and in triggering preventative and/or treatment therapies.

The above -identified and other objects are achieved by the methods of the present invention as disclosed herein which include the quantitative analysis of free L-carnitine, acylcarnitines, and total carnitine in plasma specimens dried on filter paper through suitable methods including enzymatic assays, HPLC, and electrospray MS/MS. Methods of the present invention include specimen collection, preparation and analysis through quantitative calculations and comparisons of the samples with calibration and quality control samples prepared from dialyzed plasma.

Plasma samples are collected from a dialysis patient before, during and/or after each dialysis session and spotted on filter paper and dried. The dried samples are then sent (such as via overnight mail) to a remote laboratory location for analysis. The dried specimens are then prepared for analysis at the laboratory by re constitution with an internal standard containing alcohol and then, preferably, derivatization. Prepared samples are analyzed to quantify free carnitine and acylcarnitine and their internal standards in the patient samples, calibration curve samples and quality control samples. The results of the quantification assays of the various samples are then computed to quantify free carnitine, acylcarnitine and total carnitines in the dried patient samples while accounting for, among other things, various sample collection and preparation complications, hydrolysis of acylcarnitines, interference by glutamate, and equipment performance.

Preferred embodiments of the invention analyze the samples by flow injection electrospray MS/MS using a variety of SRM and precursor ion scans. Preferably, methanol solutions which contain deuterized internal carnitine standards are used to extract the dried plasma samples from the filter paper, whether the plasma is from patient samples or known calibration curve samples or quality control samples. In most preferred embodiments, the extracted plasma samples are then followed up with gas drying and then esterification with acidified butanol, and any accompanying hydrolysis of acylcarnitines into free carnitine is accounted for in the MS/MS output data using appropriate calculations as disclosed herein.

Additional features and advantages of the invention will be set forth in the description that follows, and in part will be apparent from the description, or may be learned by practice of the invention. The invention will now be described in further detail with respect to particular embodiments thereof with reference to the figures depicted in the appended drawings. The above and other objectives and advantages of the invention will be realized and attained by the embodiments particularly pointed out in the written description and claims hereof as well as the appended drawings. The following detailed description and figures are intended to be illustrative of particular applications of the inventive concepts and are in no way to be taken as limiting to the invention as claimed.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a simplified block diagram showing the overall method according to preferred embodiments of the present invention.

FIG. 2 is a block diagram showing in more detail the steps involved in preparing the sample.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The methods will now be described in detail in relation to a preferred embodiment and implementation thereof that is exemplary in nature and descriptively specific as disclosed. As is customary, it will be understood that no limitation of the scope of the invention is thereby intended. The invention encompasses such alterations and further modifications in the illustrated implementations of the methods, and such further applications of the principles of the invention illustrated herein, as would normally occur to persons skilled in the art to which the invention relates.

The methods of the present invention involve plasma specimen collection from patients, with specimens then being spotted and dried on collection filter paper. The invention then utilizes appropriate preparation and analysis steps to produce quantitative data for comparison with similarly produced data for known calibration samples and quality control samples prepared from dialyzed plasma. Accurate quantification of the concentration of free carnitine, acylcarnitines, and total carnitine is then obtained by comparison to a calibration curve produced from data regarding the calibration curve samples. According to embodiments of the present invention, plasma samples can be collected from a dialysis patient before, during, and/or after each dialysis session for subsequent spotting on filter paper to produce a dried patient sample. The filter paper used for holding and transporting dried plasma spot samples according to the present invention can be of any suitable type, including grade 903 filter paper made available from Schleicher and Schuell, and preferably is rectangular in shape and marked so as to be divided into four quadrants. In this manner, a single filter paper can be advantageously used according to the present invention to provide multiple samples from a single patient, including two or more identical plasma spots (so as to reduce chances of contamination or destruction) or up to four different samples taken at different times (such as before, during and after dialysis) from the same patient. Preferably, the filter paper also has one or more circles printed thereon to serve as targets for the plasma spots so as to assist the specimen collector at the dialysis location in spotting the plasma on the filter paper and to assist a lab technician in identifying and recovering the plasma for later analysis. The patients' plasma spot samples, once sufficiently dried as spots on the filter paper, can be sent in sealed envelopes (such as via overnight mail) to a remote laboratory location for analysis as such dried plasma samples have been found to remain stable at room temperature for up to one week.

According to the invention, such dried plasma samples taken from dialysis patients can be analyzed according to any suitable assaying method, including enzymatic assays, HPLC, and MS/MS. According to preferred embodiments of the present invention as hereinafter described in detail, however, analysis of the plasma samples is performed by electrospray MS/MS. This assaying technique is preferred in part because it allows for the simultaneous determination of free carnitine, acylcarnitines, and total carnitine concentrations in a single analytical run.

According to such preferred embodiments of the invention, once received at the laboratory, the dried plasma specimens are prepared by re constitution with an internal standard-containing alcohol and then derivatized. Prepared samples are analyzed to quantify free carnitine, acylcarnitines and total carnitine concentrations (relative to that of the internal standards) in the patient samples, in known calibration curve samples and in known quality control samples where the calibration curve samples and quality control samples are prepared by the laboratory from dialyzed plasma. The results of the quantification assays of the various samples are then computed to quantify free carnitine, acylcarnitines and total carnitines in the dried patient samples while accounting for, among other things, various sample collection and preparation complications, hydrolysis of acylcarnitines, interference by glutamate, and mass- spectrometer performance.

In such embodiments of the present invention, dried calibration curve samples having various known compositions are prepared from dialyzed plasma and purified free L-carnitine inner salt. Five appropriate calibration curve samples comprising free L-carnitine spiked into dialyzed plasma from a methanol: water solution are itemized below in Table 1. As seen in the table, the calibration curve plasma samples have a known free carnitine concentration that corresponds to the particular free carnitine calibration curve stock solution used to p epare the plasma spot.

<img file="WO2004026114A2_D0001.tif" />

Table 1

To prepare the calibration curve solutions ("FCxxxx") indicated above, the following procedure can be used. An appropriate amount of free L- carnitine inner salt is dissolved into a 50:50 HPLC-grade methanol-water solution to make a reconstituted free carnitine concentrated calibration curve lmg/mL solution. The concentration of this concentrated calibration curve solution should be 6.203mM. Next, a first calibration curve stock solution, FC4000, is prepared by diluting 1074DL of the concentrated calibration curve solution with 3926DL of the 50:50 HPLC-grade methanol-water solution. The concentration of the free L-carnitine in solution FC4000 should be 1.333nmol/DL. Calibration curve stock solution FC2000 is then prepared by diluting 2mL of FC4000 with 2mL of the 50:50 HPLC-grade methanol-water solution. Subsequent sequential 1:1 dilutions in 50:50 HPLC-grade methanol-water solution are then used to prepare calibration curve stock solutions FCIOOO and FC0500. The 50:50 HPLC-grade methanol-water solution is used as calibration curve solution FC0000.

To prepare the calibration curve plasma samples as depicted above in Table 1, lmL of dialyzed plasma is combined with 30DL of each calibration curve stock solution FC cxxx to produce the free carnitine concentrations indicated. To prepare the dried samples, 25DL of these plasma solutions are then spotted onto the same type of filter paper used to collect the patient plasma samples and allowed to dry at room temperature (typically at least 2.5 hours). These dried calibration curve plasma samples have approximately the same stability as the patient plasma samples, and can be further preserved for later use (up to about 3 months) by storing the dried calibration curve samples at - 20°C.

Additionally in such embodiments of the present invention, dried quality control samples having various known compositions are prepared from dialyzed plasma and purified free L-carnitine inner salt and acetylcarnitine hydrochloride. Like with the preparation of calibration curve plasma samples as described above, the preparation of quality control plasma samples according to embodiments of the present invention first comprises the preparation of several carnitine -containing quality control solutions for mixing with dialyzed plasma. In preferred embodiments, these quality control solutions, all having 50:50 HPLC- grade methanol-water solution as the solvent, include a free carnitine stock solution ("FC-S"), a free carnitine working solution ("FG-W"), an acetylcarnitine stock solution ("C2-S") and an acetlycarnitine working solution ("C2-W'). Solution FC-S is prepared to have a carnitine concentration of 2nmol/DL, and solution FC-W is prepared therefrom via dilution to have a carnitine concentration of 0.5nmol/DL. Likewise, solution C2-S is prepared to have a carnitine concentration of 4nmol/DL, and solution C2-W is prepared therefrom via dilution to have a concentration of lnmol/DL.

Once the necessary quality control solutions are prepared, various quality control plasma solutions (QCyyyy) can be prepared by combining 5mL of dialyzed plasma with combinations of the quality control solutions and a pure HPLC-grade methanol:water solution as indicated below in Table 2.

<img file="WO2004026114A2_D0002.tif" />

Table 2

By way of explanation, it can be determined from Table 2 above that quality control plasma solution QC0510 is prepared by mixing lmL of FC-W and l L of C2-W into 5mL of dialyzed plasma, and quality control plasma solution QC0040 is likewise prepared by mixing lmL of C2-S and lmL of 50:50 HPLC-grade methanokwater into 5mL of dialyzed plasma. Following the procedure indicated in Table 2, nine different quality control samples QGyyyy can be produced having the free carnitine concentrations ("[FC]") and acetylcarnitine concentrations ("[C2]") indicated below in Table 3. <img file="WO2004026114A2_D0003.tif" />

Table 3

As with the calibration curve plasma samples, dried quality control samples are prepared by spotting 25DL of the quality control plasma solutions onto the same type of filter paper used to collect the patient plasma samples and allowed to dry (again, typically taking at least 2.5 hours at room temperature). These dried quality control plasma samples also have approximately the same stability as the patient plasma samples, and can be further preserved for later use (up to about 3 months) by storing at -20°C.

Referring now to FIG. 1, there is represented an overview of a method according to preferred embodiments of the present invention for quantifying carnitine concentration levels which can be utilized in diagnosing the extent of carnitine deficiency in dialysis patients within a clinical diagnostic setting. The method involves four main steps that collectively provide rapid, automated, and accurate sample analysis. Efficient sample preparation 10 of the dried plasma samples (patient plasma samples, calibration curve samples, and quality control samples are all prepared for analysis in like fashion) is necessary to insure accurate derivatization of the carnitine in the samples and constant and quantifiable hydrolysis of acetylcarnitines to free carnitine. Labeled internal standards are combined with the patient samples during their preparation to provide reference for quantification. After sample preparation 10, the samples are injected 12 into the electrospray tandem mass spectrometer. As is known in the art, an electrospray tandem mass spectrometer can be advantageously adapted to implement many automated sample-handling procedures to insure the speed and consistency of sample scanning. Data is then obtained from the MS/MS scans of the patient samples and calibration curve samples and subsequently processed to quantify carnitine levels in all of patient plasma samples in step 14. At this step, preferably values produced from the scan of the mass spectrometer are processed and printed into spreadsheet form to further allow checking of the calculations, a means of assuring accurate number production and quality. Furthermore, all spectra data is kept accurate using system diagnostic checks and quality control samples as seen in step 16. To assure diagnostic accuracy and sample quality, periodic system integrity checks and control samples that include specific additives are employed. In combination, the above-mentioned steps maximize the rate and quality at which plasma samples are analyzed for carnitine levels to result in a method that is suitable for use in the clinical setting.

FIG. 2 shows an overview of the sample preparation procedure (step 10 of FIG. 1) used in preferred embodiments of the present invention. An initial sample login 20 is performed by coding each sample, thereby associating the sample to a specific location of a microtiter well in a microtiter plate. The samples consist of plasma spots placed on designated areas of filter paper. After each spot is punched 22, such as into disks with an automated puncher (such disks typically having a diameter in the range of 3/16in to l/8in, and preferably 3/16in), it is placed into the designated microtiter well of a flat-bottom microtiter plate. Internal standard preparations, which were prepared with methanol or some other suitable alcohol that serves as an extraction solvent, are then added 24 to the dry punched plasma spot in each well. Such additions 24 are preferably performed using automated sample handling equipment.

The methanol, or other suitable alcohol employed, in the internal standards preparations serves as the solvent extraction solvent while the internal standards serve to quantify the free carnitine and acylcarnitines in the dried plasma matrix of each sample. In embodiments of the invention, the internal standard preparations are adapted to comprise an ideal mix of acylcarnitine/carnitine deuterium-labeled standards dissolved in an extraction solvent. As will be understood by one skilled in the art, the internal standards are provided in the extraction medium for the optimum mixed-mode scan functions to maximize carnitine detection. A list of the free carnitine and acylcarnitine internal standards used in preferred embodiments of the invention are listed below in Table 4. A suitable extraction solvent containing labeled internal standards can be prepared according to the following procedure. A set of isotopically labeled L-carnitine and L-acylcarnitine standards, such as in solid form, can be commercially obtained so as to facilitate the preparation of mixtures wherein the stoichiometric ratios of the carnitines are known. According to preferred embodiments of the present invention, the labeled L-carnitine and L- acylcarnitines include [Dg] carnitine, [D₃] acetylcarnitine, [D3]propionylcarnitine, [D3]butyrylcarnitine, [D9]isovalerylcarnitine, [D3]octanoylcarnitine, [Dgjmyristoylcarnitine and [D₃]palmitoylcarnitine and are present in the internal standard in the molar ratio 20:5:1:1:1:1: 1:2 as depicted in Table 4 below.

<img file="WO2004026114A2_D0004.tif" />

Table 4

A suitable internal standard working solution for analyzing individual 3/16in punched plasma spot disks, such as solution "IS-Working" in

Table 4 above, can be prepared according to the following procedure. The various labeled L-carnitine and L-acylcarnitines are reconstituted in HPLC-grade methanol (or other suitable extraction solvent alcohol) to form a reconstituted internal standard solution having the solute standard concentrations of solution IS -Reconstituted in Table 4. A concentrated stock internal standard solution ("IS-Stock" in Table 4) is then prepared by diluting 500 DL of solution IS-

Reconstituted in 3500 DL of an HPLC-grade 50:50 methanol:water solution to achieve a 1:8 dilution of the internal standards. Finally, a working extraction solvent solution ("IS-Working") containing the labeled internal standards for use in recovering the plasma from the dried samples is prepared by making a 1:150 dilution of solution IS-Stock in HPLC grade methanol. IS-Working is used in preparing the dried plasma samples for tandem mass spectrometry as will be described below.

The concentrations of the internal standards in IS-Working solution can, of course, be readily adjusted to analyze two 3/16in or two l/8in dried plasma spots, a single 3/16in or l/8in dried plasma spot, etc., by adjusting the volume of the extraction solvent additions to the analyzed samples or the concentration of the IS-stock solution in a manner as will be apparent to one skilled in the art. The working standard solution IS-Working thereby serves as both the extraction solvent and the means for internal standardization of the analysis. In these preferred embodiments of the invention, each analytical run of the MS/MS includes a combination of calibration curve dried plasma samples and quality control dried plasma samples along with the patient specimen samples which one desires to have analyzed for carnitine concentrations. For example, all five different calibration curve plasma samples (FC4000, FC2000, FCIOOO, FC0500, FC0000) and any two randomly selected quality control samples (such as QC0000 and QC2040, or QC2010 and QC0040) can be interspersed with patient plasma samples on each microtiter plate introduced into the electrospray mass spectrometer for analysis. This procedure helps establish a calibration curve for analyzing the patient plasma samples and methods to account for various inconsistencies and errors produced in the MS/MS preparation and analysis. All three types of dried plasma samples are prepared for analysis according to the same procedure indicated in FIG. 2. Referring again to FIG. 2, in preferred embodiments of the invention, dried calibration curve plasma samples, dried quality control plasma samples and dried patient plasma samples are first punched into recorded wells of a flat bottom microtiter plates, such as with a Wallac DBS puncher fitted with a 3/16in punch head, at steps 20 and 22. Next, at step 24, an appropriate amount, such as 300 DL, of an internal standard-containing working solution (e.g., IS-Working) is added to each sample in the plate. This solvent addition can be advantageously automated by using a Gilson liquid handling system. The microtiter plate containing the mixture of all three types of samples is then agitated on rotator for approximately 30 minutes to extract plasma 25 from the dried plasma sample disks.

The extract from each sample is then transferred from the flat bottom microtiter plate to a round bottom plate, such as via the Gilson liquid handling system, to allow for further processing. The solvent from the internal standard is then removed 26 via evaporation, preferably under a gentle stream of nitrogen using a SPE-Dry nitrogen evaporator. In embodiments of the invention wherein samples are derivatized, each sample in the microtiter plate then undergo esterification 27. This is done by first adding an appropriate amount of acidified alcohol, such as via an Eppendorf pipette. Preferably, the derivatizing agent comprises 50 DL of 3mol/L HCl in n-butanol to each sample well. In the esterification step, the samples in the microtiter plate are then covered with blue septa sheet fitted to the size of the plate, and then with a metal plate before the plate is placed in an oven and heated to stimulate derivatization. Preferably, the heating occurs at approximately 65°C for about 15 minutes. The derivatized specimens are then immediately dried 28 in the nitrogen evaporator as before to remove the volatile derivatization agent. After this drying step to remove derivatization agent, the samples are then reconstituted 29 by adding mobile phase, typically about lOODL, by means of the Gilson liquid handler. The microtiter plate is then sealed with a clear micromat plate cover before analysis to retard evaporation of the mobile phase.

As described herein, preferred embodiments of the present invention employ derivatization of the carnitines in the plasma samples in order to increase sensitivity of the assay. Optionally, of course, the derivatization step can be omitted from the assay when appropriate, such as when relatively larger carnitine concentrations are being quantified.

Before sample analysis, optimization of the MS/MS systems is achieved by regular injection of a tuning solution, such as daily. In preferred embodiments, the sensitivity threshold is determined from scans of [Dsjphenylalanine which should have a sensitivity of about at least 25% with a maximum intensity set to 5xl0⁶ counts per second ("cps"). The electrospray MS/MS system employed according to these preferred embodiments is a low flow rate system employing the use of a fused silica line displaced to the tip of the electrode. Automated injection systems use the fused silica line to directly connect the injector to electrode tip to minimize dead space. The mobile phase flow rate of the MS/MS can be set to about 18.0DL/min and the injection volume can be set to about 15DL to achieve reliable results with about a 2 minute cycle time. The scans implemented to detect the necessary fragments of the ions include positive SRM and positive precursor scans. Suitable instrument parameters, which may be adapted as is known in the art for optimization, are depicted below in Table 5.

<img file="WO2004026114A2_D0005.tif" />

Table 5 Furthermore, an example of suitable general scan parameters are provided below in Table 6.

<img file="WO2004026114A2_D0006.tif" />

Table 6

An example of suitable scan parameters specifically for the positive SRM scans are listed below in Table 7.

<img file="WO2004026114A2_D0007.tif" />

Table 7

Similarly, an example of suitable scan parameters specifically for the positive precursor scans (for precursor of 85.05amu) are listed below in Table 8.

<img file="WO2004026114A2_D0008.tif" />

Table 8

After the MS/MS scans are preformed according to the present invention, the data obtained must be analyzed to calculate free carnitine, acylcarnitines, and total carnitine concentrations present in the patient samples. Preferably, these calculations are automated using a computer and appropriate MS/MS data acquisition and commercial data manipulation software. A suitable data acquisition software includes the Chemoview software and Analyst software

49 programs made available from Applied Biosystems. Specifically, the Chemoview software enables automated calculation of the relative ratios of an analyte (e.g. free L-carnitine) to its internal- standard (i.e., [Dg]L-carnitine) by allowing the user to define analyte masses of interest and associate relevant internal standard masses and desired internal standard concentration therewith. The data acquisition software also multiplies the analyte-to-standard mass ratio by the concentration of the internal standard to quantify the concentration of the analyte. Unfortunately, the hydrolization of acylcarnitines into free carnitine and the relative amounts of internal standard to analyte requires corrections to the analyte concentrations provided by the data acquisition software. Preferably, these corrections are thereafter performed by the data manipulation software (such as as, for example, a commercially available spreadsheet program).

First, a post-acquisition data correction must be made to adjust for the relative concentrations of the unknown analytes in the analyzed plasma samples by estimating the volume of plasma typically recovered from the punched disk (the amount and concentration of the internal standards being known). Assuming 25DL of plasma were spotted on the filter paper, empirical observation concludes that a 25DL spot on suitable filter paper produces a spot having an average diameter of 13.4mm, or equivalently a radius (assuming a circular shape) of 6.7mm (or 0.264in). The average surface area of a 25DL plasma spot on filter paper then would be about 0.219in² (calculated using the standard equation for the area of a circle). One could also calculate that a 3/16in diameter punch would produce a disk having a surface area of 0.0276in² while a l/8in punch would produce a disk having a surface area of 0.0123in². Thus, it is apparent that all of the plasma sample spotted on the filter paper will not be recovered by either a l/8in or 3/16in punch. Therefore a smaller volume of plasma (and the carnitine it contains) than the full 25DL will be reconstituted. To account for this factor, one can estimate the volume contained in a punch according to the following equation i whereby:

(equation i) Vol. = 25DL*(Areaf κA / AreaAυerage piasm Spot) Thus, where a 3/16in punch is used, one can estimate that approximately 3.2DL of plasma is contained in a disk punched from the dried plasma samples.

This average volume of plasma contained in a punched disk can then be used to obtain a conversion factor for each analyte to apply against the concentrations obtained from the data acquisition software. According to the procedure outlined above, for [Dg] L-carnitine about 300DL of IS-W is added to each punched plasma disk, which means that approximately 38pmol total of [Dg]L-camitine is mixed with the 3.2uL of plasma that is reconstituted from the 3/16in disk. Thus, the concentration of [Dg]L-carnitine in the analyzed plasma for each sample is known to be approximately 11.9 Dmol/L. Data acquisition softwares for use with tandem mass spectrometers typically provide either relative concentrations of each analyte with respect to their corresponding internal standards or alternatively provide baselined absolute concentrations of each analyte by assuming a baseline concentration for its internal standard. The conversion factor therefore allows the calculation of an accurate absolute concentration of each analyte. For example, since Chemoview assumes a nominal internal standard concentration of lOmmol/L, the conversion factor to multiply the L-carnitine concentration data is 1.19 (11.9Dmol/L actual / lO.OODmol/L nominal). Similar calculations are applied to each analyte to obtain similar conversion factors and internal standard volume-adjusted data.

The data from the known calibration curve samples is used by the computer to interpolate the concentration values of carnitine in the patient and quality control samples by the construction of calibration curves. The calibration curves, or standard curves, produced by the data acquisition software allows for interpolated concentrations to be taken as the true concentrations in a manner as is known in the art.

This volume-adjusted raw data (comprising MS/MS volume adjusted concentrations of analytes) is then exported to a data manipulation software package, such as a spreadsheet program, to apply equations ii through v below to correct for glutamic acid interference and hydrolysis of acylcarnitines into free carnitine. Analyses using MS/MS must account for the fact that mass spectrometers cannot readily distinguish analytes with the same molecular masses that also have the same product ions and molecular ions (such as isotopes). An example of such problematic analytes are leucine and isoleucine. However, it is known that if no product ions are shared, then ions of the same molecular mass do not interfere in MS/MS analyses. Although the butyl ester of glutamic acid and acetylcarnitine do not have the same molecular masses (259 m/z and 260 m/z, respectively), the protonated form of glutamic acid shares the same molecular ion mass of acetylcarnitine (260 m/z). Glutamate has physiological concentrations 10 - 100 times greater than acetylcarnitine.

Therefore, the signal at 260 m/z, used for the detection of acetylcarnitine, is also affected by the glutamic acid contribution. However, the natural isotopic abundance of glutamic acid, and particularly its protonated form, could be used for estimating the contribution of glutamic acid to the signal for the detection of acetylcarnitine. According to embodiments of the invention, the contribution of glutamate at 260 is calculated by referencing the signal at 261 since 261 is also the M + 1 natural isotopic abundance of acetylcarnitine (sometimes abbreviated herein as "C2"). A corrective formula is employed that first estimates the isotopic contribution at 261 by acetylcarnitine. The difference between the estimated contribution at 261 by acetylcarnitine and the actual raw data represents the contribution from glutamate. Finally, using the doublet of nearly equal intensity for glutamate at 260, the actual concentration of acetlycarnitine can be determined.

To get a corrected acetylcarnitine concentration that accounts for glutamic acid interference (i.e., "G2corrected"), equation ii below is employed, whereby:

equatWTl 11) G2Corrected — G2Measured — G2Glutamate

wherein C2Measured is the volume-adjusted actual acetylcarnitine concentration value provided by the data acquisition software, and G2Gi tamate is the actual glutamate contribution. C2 Glutamate can be determined from the following equation Hi, whereby:

(equation Hi) G2 Glutamate = Glutamate_?6i - C2z_so

wherein Glutamate≤βi is the volume-adjusted actual glutamate concentration value at 261 provided by the data acquisition software, and C2j_so is the estimated isotope concentration at 261 of acetylcarnitine. C2/so in turn can be estimated from the following equation iv, whereby:

(equatio W) G2lso = 0.14 X G2Measuredat 260

wherein the value of factor 0.14 is the determined M + 1 natural abundance contribution of acetylcarnitine, and G2Meas red at 2βo is the volume -adjusted actual acetlycarnitine concentration value at 260 provided by the data acquisition software.

To obtain corrected values of free carnitine concentration in a patient sample that accounts for acylcarnitine hydrolysis, equation υ is employed, whereby:

equatio ϋ) FGCorrected = FGMeasured — FGHydrolyzed

wherein FGMe sured is the volume-adjusted actual free carnitine concentration value for a given sample as provided by the data acquisition software, and FGHydroiyzed is the portion of FGMeas ed that is due to the hydrolysis of acylcarnitines into free carnitine. FGnydroiyzed in turn can be calculated from the following equation vi, whereby:

equatio Vi) FC Hydrolyzed = %Hydrolysis X AC

wherein %Hydmiysis is the percent of measured free carnitine that is produced from hydrolysis of acylcarnitines, and AC is the volume-adjusted actual total acylcarnitine concentration value taking into account the corrected value for the concentration of acetylcarnitine. Given that the hydrolysis of acylcarnitines occurs at essentially a constant rate, the %H_ydroiy is can be calculated from the known internal standard concentrations by employing the following equation υii, whereby:

(equatio VU) VoHydrolysis = <img file="WO2004026114A2_D0009.tif" /> I 5.0DM

wherein <img file="WO2004026114A2_D0010.tif" /> is the deuterated free carnitine concentration measured at 221 and the value 5.0DM represents the concentration of deuterated acetylcarnitine internal standard utilized in the samples.

To calculate the actual total acylcarnitine concentrations ("AC"), equation υiii is employed, whereby:

(equation υiii) AC = <img file="WO2004026114A2_D0011.tif" /> C3, C4, C5:l, C5,

C40H, C6, C5OH, C60H, C8:l, C8, C8OH(C3DC), C10:2, 010:1, CIO, C4DC, C5DC(C10OH), Cl2:l, C12, C6DC, C12OH, C14:2, C14:1, C14,

C140H, C16:1, C16, C16OH, C18:2, C18:l, C18)

Finally, total carnitine concentration can be calculated from the corrected free and acylcarnitine concentrations according to equation ix below.

(equation ix) TC = FGcorrected + AC

The designations of the acylcarnitine butyl esters indicated above in equation viii can be recognized using the following

Table 9 ("*" indicating an internal standard).

<img file="WO2004026114A2_D0012.tif" />

Table 9 As will be appreciated by one skilled in the art, the above carnitine level quantification methods according to embodiments of the present invention can be advantageously utilized in diagnosing the extent of carnitine deficiency in dialysis patients within a clinical diagnostic setting. Such diagnosis could in turn be used by clinicians to select and implement appropriate preventative and treatment therapies including the intravenous administration of injectable levocarnitine following dialysis sessions for ESRD patients. It is possible that, in a clinical diagnostic setting, the above methods and constituent calculations may be interpreted so a proper course of action can be taken depending on the concentration levels of carnitine.

In order to facilitate the diagnosis of patients, decision trees could be used to interpret the level of the carnitine deficiency, and which is used to assist the user or interpreter in determining the next course of action and the significance of the concentration reading. While either the initial concentration values or the corrected values may be applied to the interpretation guide, it is, of course, preferred that the corrected values for acylcarnitine and free carnitine are used because of the heightened accuracy.

Various modifications of the embodiments herein disclosed will be readily apparent to one skilled in the art after reading the above. For example, any known suitable sample analysis assay can be used in alternative embodiments of the invention to determine the carnitine concentrations in the dried plasma samples. It will be understood that the above MS/MS methods and their concepts could be adapted to the analysis of other biological specimens, including urine, blood, spinal fluid, and tissue extracts (e.g., homogenized with an appropriate liquid vehicle and centrifuged). Any and all such modifications are intended to be covered by the application as claimed.