Test element and method for quantitative NIR spectroscopic analysis

IS 2 155 167 T3

DESCRIPTION

Test item and procedure for quantitative NlR spectroscopic analysis.

This invention relates to the assay of analytes in biological liquids using NIR near infrared spectroscopy and a dry slide assay element.

For the past 11 years or more, the standard dry reagent procedure for clone cheomics (not using liquid reagents or reconstituted liquid in a cuvette) has been to use dry slide test item technology such as that used by the Eastman Kodak Company under the trademark "Ektachem" slides. In such technology, a dry reagent is supplied together with a binder in a reagent layer on a slide-like element. To help spread the liquid evenly in an area which then flows into the reagent layer, a spreading layer arranged above the reagent layer is usually provided, such as, for example, that outlined in US Pat. A-3,992,158. However, the elements are read by reflectance from the opposite side, and all efforts have been directed to prevent the detection light from penetrating the spreading layer, since the detectable color change takes place in the ( s) reagent layer (s).

Although such test items perform admirably, as demonstrated by the billions of slide items sold since Eastman Kodak introduced the product, it would be advantageous to provide a dry test item that does not require the various reagents needed in, for example, the aforementioned patent 3,992,158 and related test items.

Some efforts have been made to transform the spectroscopic analysis of liquids into a quantitative science. An attempt was made to use waterproof supports on which a liquid sample was spread. However, for the quantitative analysis of most biological liquids this is not satisfactory, since in an impermeable support the constancy of the light path required by Beer's law cannot easily be maintained. There is a similar objection to earlier permeable supports that were attempted to use, eg, paper and thin-layer chromatography elements. That is, although in such permeable supports it is easier to maintain the depth of the examined liquid, they lack a uniform three-dimensional porosity, since the liquid does not spread uniformly in all directions along a known volume.

The most recent attempt in spectroscopic analysis is the use of microporous polyomers to create a porous lamine for IR spectroscopy, for example, as taught in WO 93/00580. Although it is intended to provide detection wavelengths that cover the NIR spectrum, as well as the IR (p. 1, lines 25-28), with NIR as used in this document being 750 nm to 3000 nm, really the rest of the teachings are only refers to the IR spectrum (above 3000 nm). The reason is that, in the literature, it is known that the laminated materials indicated in WO 93/00580 cannot work quantitatively in the NIR, particularly in the reflection mode in which the lamina has to have a reflection of at least 95%. All listed materials (polyethylene, polypropylene, poly (tetrafluoroethylene) (PTFE), ethylene / propylene copolymers, poly (vinyl fluoride), polyester, chlorotrifluoroethylene polymer and nylon), with the exception of PTFE, lack a they reflected at least 95%. However, PTFE does not have the required porosity - it is not "porous enough" as noted later in the Summary.

The present inventors have discovered a microporous material that avoids the problems noted above.

More specifically, according to one aspect of the invention, the problems are solved by a test element for analyzing analytes present in patient samples, which comprises a support, a layer that provides a substantially constant light path comprising a material of diffuse reflection that a) is a reddish-colored polymer, microcrystalline particles or adhered particles, with or without reflection pigments, b) it is sufficiently porous in all directions to allow a liquid to spread uniformly in all directions, as measured by the test indicated in the paragraph titled "Porosity Test" in the present description and c) it reflects homogeneously at least 95% of NIR radiation, the element substantially lacking reagents capable of reacting with the analyte.

The test element may also include a hydrophilic polymer layer, separated from the layer that provides the light path and with a thickness of less than 1 micrometer.

The test element may also include a mirror layer between the layer that provides the light path and the support, effective to increase the length of the path of NIR radiation through the element.

In accordance with another aspect of the invention, a method is provided for quantitatively analyzing an analyte in a biological sample by means of near infrared (NIR) scattering or reflection spectroscopic quantitative anaolysis, comprising the steps of:

a) placing an aliquot of sample on a test element according to the first aspect of the invention;

b) illuminating the sample within the element with light of wavelengths of the NIR spectrum;

c) detecting light of such wavelength that it is reflected or scattered by the element; Y

IS 2 155 167 T3

d) spectroscopically analyzing the detected light with respect to the signals that are characteristic of the analyte being analyzed.

In the above procedure, step d) may comprise comparing the signal with a calibrated model for the analyte of choice based on the initial detected signal or derivatives thereof, to find out the concentration of the analyte of choice from the calibrated model. .

The present invention also provides the use of an assay element according to a first aspect of the invention, to analyze analytes in patient samples by means of a near infrared reflection or scattering spectroscopic quantitative analysis.

Fig. 1 is a schematic sectional view of a prior art assay element with dry reagents, and the procedure used to analyze the analyte of choice;

Fig. 2 is a schematic view similar to that of Fig. 1, but of the invention;

Fig. 3 is a partially schematic plan view of a useful spectroscope;

Figs. 4-7 are graphs of actual analyte concentrations versus predicted concentration using the invention; and Fig. 8 is the gelatin absorption plot at NIR wavelengths, demonstrating why gelatin is unacceptable at significant thicknesses in the layers defining the light path of the test element.

The invention is based on the discovery that the particular extension layers described in the aforementioned patent 3,992,158 are very suitable for quantitative NIR spectroscopic anaolysis when they comprise the layer that defines the path of light, as opposed to what occurs in the construction of patent 3,992,158.

Referring to Fig. 1, the conventional slide member 10 taught by the '1958 patent comprises a transparent plastic support 15, one or more layers 20 of reagent, and a layer 30 of uniformly porous extension. Layer 30 provides the uniform spread of a sample drop (shown in transparency) in a volume of area A dispersed in layer 30 (shown in liquid), which then migrates into the reagent layer 20. (Layer 20 reagents are indicated by thicker stippling). Importantly, the element is read from the opposite side by reflectance, using a light source 40 and a detector 42 in a reflectometer 50. The extension layer, in turn, has a screen material that makes it reflective, such as by incorporation of TiO2. As a result, the light path 60 is ideally defined only by layers 15 and 20 and not by the extension layer, so that a color change is detected in layer 20. (Although in some examples of the “Ektachem” ® brand slides manufactured in accordance with the teachings of the patent 3,992,158 some of the light can indeed be transmitted into the extension layer, further on it is shown that many slides “ Ektachem ”in fact have less than 1% of the light path operationally defined by that layer, that is, within the normal operating range of the slide. In any case, such slides have always been taught in connection with use with detection reagents.)

The present invention is based on the discovery that certain particular spreading layers themselves constitute an excellent porous lamin for performing quantitative NIR spectroscopic analysis, when the light path is effectively defined by the spreading layer rather than by any layer. adjacent. The reason is that the sufficiently porous nature of the spreading layer makes it ideal for providing a constant predictable light path length over the period of use, unlike the binders used in the reagent layers of Fig. 1 that define the light path within the slide element.

As used herein, "NIR" spectroscopy or radiation means the use of the wavelengths from 750 to 3000 nm. "Substantially constant light path" means with deviations not greater than 1%. "Quantitative" means the measurement of actual quantities based on the accuracy and precision that they are acceptable in a clone laboratory.

"Sufficiently porous" refers to all directions so that the liquid spreads within the layer evenly in all directions. As indicated, the latter is a property that is already known to exist in the spreading layer of the element taught in patent 3,992,158. Therefore, "sufficient porosity", as used herein, refers to any layer of material that is as porous in all directions as the extension layers of the patent 3,992,158. The TiO2 layer from Ex. 1 provided below is a successful example of this.

To test materials having adequate porosity, the procedure outlined in column 6, line 22 to column 7, line 7 of the aforementioned patent 3,992,158 is used. This test is described in the following paragraph under the heading "Porosity Test". However, it is not necessary to follow the volumetric and thickness requirements stated in this test nor does a gelatin undercoat have to be used. Alternatively gelatin can be used as it is a porosity test.

Porosity Test

To perform such a test, a transparent photographic film backing material, such as diffuse polyethylene terephthalate, can be applied to a transparent gelatin layer with a gelatin coverage of approximately 200 mg / dm.². Gelatin can vary in hardness but a layer of hardened gelatin that increases the thickness of

ES 2 155 167 T3 the layer by about 300 percent when soaked for 5 minutes in 22 ° C water. When dry, the gelatin layer will be approximately 30 microns thick. On the gelatin layer, the extensioan layer to be evaluated can be applied, such as by coating with a solution or dispersion. The extensian layers can be designed to have widely varying dry thicknesses and, for testing purposes, a thickness of from about 100 to about 200 microns is desirable. After the layers have dried, a sample of test solution or dispersion may be applied to the surface of the extension layer under evaluation, preferably in a small amount so that not all portions of the layer are moistened by the applied sample, but desirably sufficient to create a moistened region that may include a circular area with a diameter of approximately 8-10 millimeters. The selection of a test solution or dispersion should be chosen and will depend in part on the type of sample or analyte to which the coating will be exposed under actual use conditions. For low molecular weight materials, aqueous dye solutions can be used and a 0.0005 weight percent solution of Solatine Pink is acceptable. For higher molecular weight materials, such as proteans, an aqueous dispersion of bovine albumin colored with Solatine Pink can be used. After applying the liquid sample to the layer under evaluation and allowing the liquid sample to disappear from the surface and collect on the layer, the test item can be turned over so that the lower surface of the spreading layer can be visualized. Proposed through the transparent support material and the gelatin layer. If prior to substantial evaporation of the solvent or dispersion medium the colored spot has a substantially uniform color density when examined with a densitometer having a circular aperture of approximately 2-3 millimeters, then the spread and the achievement of of a uniform apparent concentration on the lower surface of the test layer and / or on the gelatin layer, and the test layer may be useful as an extensioan layer in analytical elements of the type described herein. By "substantially uniform density" is meant a density across the spot, with the exception of its periphery, which has a maximum and minimum value no greater than ± 10-15 percent of the mean density. Due to edge effects, uncharacteristic density gradients may appear at the periphery of the spot, but they do not necessarily have an effect on the meaning of an analytical result. Peripheral area may vary between spots, but will normally not be greater than about 20 percent of the entire spot and may be less.

Thus, it will be evident that the basic components of this invention are conventional: the components and the construction of the layer (s) that define (s) the light path of the test element and the NIR analysis for detection. signal peaks representative of certain analytes, for example, total proteans, glucose, total cholesterol, albumin, globulin, triglycerides, urea, creatinine, HDL, and LDL.

The invention resides in the discovery that certain particular assay elements of the patent 3,992,158, when properly constructed substantially without reagents and without the reagent layer, will allow adequate quantitative detection of analytes in biological fluids using NIR spectroscopy. For these reasons, except where noted herein, no details are provided for the construction of the test item or the performance of the NIR analyzes, as such details are well known to those of ordinary skill in the respective art. (Representative useful peaks for, eg, total protein, include 2050, 2160, 2170, and 2180 nm, as is well known).

Although the description presented herein refers to preferred embodiments that characterize reflection NIR assays using particular extension layers of US-3,992,158, the invention is not limited thereto. It is also useful with NIR dispersion-type analysis.

The preferred slide test element is element 100, FIG. 2. This comprises a layer 30 'which is substantially the same as the particular layers of the extensioan layers of the aforementioned patent 3,992,158, provided it lacks the reagents of detection. Optionally, it is attached to a NIR light transparent plastic backing 15 by a conventional bottom layer 41, or by corona or plasma treatment of the backing. The supports described in the patent 3,992,158 are useful. "Thin", as used herein for the bottom layer, refers to a thickness less than 1 micrometer, as a greater thickness tends to interfere with NIR reflection, especially if the bottom layer comprises a hydrophilic polymer.

Alternatively, if the composition of layer 30 'is self-adhering to layer 15, layer 41 can be omitted.

Detection of analytes in a liquid deposited in area A is performed by a conventional spectrophotometer 200, using an appropriate NIR energy source 210 and a NIR detector 220. (Although in a preferred mode of analysis it is done from the front side as shown, it can be done from the back side, shown in transparency).

Materials for use in layer 30 'of element 100 are selected from the group consisting of a reddish colored polymer, microcrystalline particles and adhered particles, all with or without reflective pigments, as known from the patent 3,992,158.

When performing the upper side NIR analysis of the test element 100, the layer 41 optionally includes a mirror surface which may be provided in the form of a metallized layer, a fine particle ceramic layer, a fine metal oxide, or a metal dispersed in a transparent polymer. This increases the reflection of path 300, since without the mirror surface, the NIR energy from source 210 is more likely to be partially absorbed.

ES 2 155 167 T3 at or near layers 41 and 15.

As indicated, NIR analysis is conventional. In summary, it is done as follows:

A slide element is attached to a support and placed in the position shown as "S" in Fig. 3. A spectrum of the slide is collected using a conventional 400 spectrometer and the wavelengths generated by the light source 402, monochromator. 404, monochromator graticule 406, focusing lenses 408, and detectors 410. Wavelengths are selected from the entire spectrum of 1200 to 2400 nm, to achieve a white reading. The element is then removed from the spectrometer and drops of the liquid to be tested are applied in a preferred amount such as 300 microliters. The liquid is allowed to spread into the slide for 30 to 90 seconds and the slide is then returned to the spectrometer. A new spectrum of the slide and liquid is generated and collected using the 400 spectrometer. The results are subsequently analyzed using the second derivative of absorbance followed by multiple linear regression (MLR) to distinguish the analyte from the background effect. The following examples show more details.

Examples

The following examples are illustrative only and do not constitute an exhaustive list of embodiments of the invention. In the following examples, the slide test item had the following construction:

(Top layer)

Uncle₂ or BaSO₄Cellulose Acetate (Bottom Layer)

Binder = poly (vinyl pyrrolidone) (Coverage = less or = 1 mg / cm²Poly (ethylene terephthalate) support

The protean examples listed below were calibrated and read with 2057.5 nm and 2180.5 nm light. In addition, however, the 20502250 nm and / or 1500-1800 nm regions are also useful. The spectroscope used was a “Quantum 1200 Plus” (TM) spectroscope available from LT Industries, shown schematically in Fig. 3. To distinguish the analyte from the background effect, the second derivative of the absorbance will be used followed by a multiple linear regression (MLR). ). The second derivative is preferred because it eliminates non-chemical differences, such as initial offsets and instrument interferences, and increases the overhangs and inflections of the starting absorption spectrum curves. See, for example, US-A-5,197,470 for an example of the procedure used herein. Alternatively, multiple linear regression analysis may be substituted for principal component and / or least squares regression and / or partial analysis.

Example 1

Total Serum Protein on a TiO2 Slide

In the slide construction shown above using a TiO2 matrix with a thickness of 50.8-152.4 μm at a coverage of 1-6 g / m², 300 μ! of a calibrating fluid to which known amounts of total protein had been added which are indicated on the horizontal axis of Fig. 4. The spectroscopic readings were processed so that the MLR predictions are plotted, Fig. 4, on the vertical axis and the a priori real values ​​are represented on the horizontal axis, producing a value r, where “r” is the regression correlation value, of 0.997 and a tactic deviation of 0.11. Clearly, this result is well within the laboratory standards of a quantitative assay.

These results also constitute the calibration curve used in subsequent examples. Example 2

Total Serum Protein on a TiO2 Slide

The experiment of Example 1 will be repeated, except that the calibration curve of Fig. 4, Example 1 will be used to test other known analyte concentrations not used in the preparation of the calibration curve of Example 1. The Results are represented in Fig. 5. The “r” value was 0.996 and the tactic deviation was 0.12, indicating that the results of Ex. 1 could be used as a calibration curve with “unknown substances”.

Example 3

Total Cholesterol on a BaSO4 Slide Assay Element

The experiment of Example 1 was repeated with the exception that a layer defining a light path (other than the support) was used composed of BaSO4 at a coverage of 1-6 g / cm², instead of TiO2. The cholesterol levels indicated on the horizontal axis of Fig. 6 had been added to the calibrators. The predicted values ​​obtained spectroscopically were those represented on the vertical axis, producing a result of r = 0.99 and a normal deviation of 9 mg / dL, which are within acceptable limits for this range of cholesterol concentrations. Example 4

Total Cholesterol on a BaSO4 Slide Assay Element

Example 2 will be repeated with the exception that the assay was performed for total cholesterol, using the calibration curve generated in Example 3. The results appear in the graph of Fig. 7, with values ​​of r and a tactic deviation that are identical. to those of Fig. 6.

Comparative Example No. 1

Fig. 8 is included in alone to demonstrate that any substantial thickness [greater than 1 micrometer] in the slide test element comprising gelatin will not provide a NIR radiation reflection of at least 95%. Of course, the reason is due to the substantial absorption that occurs particularly at 1500 to 2400 nm, a significant portion of the NIR range of the invention. (In this respect it has no im5

ES 2 155 167 T3 bearing the type of gelatin). Therefore, gelatin should be avoided, at least in thicknesses> 1 micron.

Comparative Example No. 2

To demonstrate that the conventional "Ektachem" (trademark) slide for eg glucose does not allow light from the reflectoometer to pass into the spreading layer for the analyte ranges for which the slide is intended to be used, He tested such a glucose slide on a standard Perkin Elmer "Lambda Nine" (trade mark) spectrometer with a known glucose concentration. The peak absorbance of the dye was 1.22 absorbance units, which indicates a transmission of approximately 7%, such that the light contributed by reflectance within the spreading layer (above the colored layer) is 0, 07 x 0.07 = 0.0049, or just 0.3% to 0.5% of the signal, instead of the required value of at least 95%. That is, the conventional "Ektachem" slide is not designed to transmit light into and through the spreading layer at the intended analyte concentration ranges.