METHOD FOR CONSTRUCTING QUADRANTS WITH MULTIPLE INDEPENDENT BIOMARKERS FOR DIAGNOSING DISEASES

Abstract

The present invention relates to a method for constructing quadrants corresponding to different diseases in a frame of concentrations of multiple independent biomarkers, comprising: (a) transferring original distributed concentrations of every independent biomarker to modified distributed concentrations, comprising: calculating the mean value and the standard deviation of the original distributed concentrations for a given independent biomarker;individually subtracting all the original distributed concentrations by the mean value for the given independent biomarker; andindividually dividing all the differences by the standard deviation to get the modified distributed concentrations for the given independent biomarker;(b) positing the modified distributed concentrations in a frame of multiple independent biomarkers; and(c) finding a boundary optimally separating neighboring quadrants corresponding to different diseases.

Description

FIELD OF THE INVENTION

The present invention relates to a method for constructing quadrants with multiple independent biomarkers for diagnosing diseases.

BACKGROUND OF THE INVENTION

Neurodegenerative diseases include several kinds of pathologies causing versatile diseases, such as Alzheimer's disease (AD), Parkinson's disease (PD), dementia with Lewy body (DLB), and frontotemporal dementia (FTD), etc. Due to different causes, these diseases may develop various impaired clearance of biomolecules, which are regarded as biomarkers with these diseases. For example, β-amyloid (Aβ), tau protein and their derivates are typically related to AD, while α-synuclein and its derivates are representative biomarkers for PD and DLB. It is suggested to discriminate patients by assaying the typical biomarkers. For example, AD patients show higher values for the product in concentrations of plasma Aβ_1-42 and tau protein, denoted as φ_Aβ1-42×φ_tau, as compared to that of normal controls (NC, or referred as to healthy subjects) and DLB patients, as shown in FIG. 1A. But, in FIG. 1A, DLB patients will be miss-diagnosed as normal controls. As the plasma α-synuclein is assayed for NC, DLB patients, and AD patients, patients with DLB show higher concentrations of plasma α-synuclein, denoted as φ_α-syn, as compared to NC, as shown in FIG. 1B. However, some AD patients also show the same levels of plasma α-synuclein as DLB patients. Thus, DLB patients can not be well discriminated from AD patients by merely assaying φ_α-syn. According to results in FIGS. 1A and 1B, DLB patients can not be accurately diagnosed by assaying either of plasma φ_Aβ1-42×φ_tau or plasma φ_α-syn.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1A: Detected concentration products of plasma Aβ_1-42 and tau protein for normal controls (NC) and patients with either dementia with Lewy Body (DLB) or Alzheimer's disease (AD). FIG. 1B: Detected concentrations of plasma α-synuclein for normal controls (NC) and patients with either dementia with Lewy Body (DLB) or Alzheimer's disease (AD).

FIG. 2: A proposed flow chart to discriminate subjects belonging to NC, AD, or DLB group by assaying multiple bio-markers.

FIG. 3A: Detected concentrations of Aβ_1-42 in plasma with variously neurodegenerative diseases. FIG. 3B: Detected concentrations of tau protein in plasma with variously neurodegenerative diseases. FIG. 3C: Detected concentrations of α-synuclein in plasma with variously neurodegenerative diseases.

FIG. 4: Product in concentrations of plasma Aβ_1-42 and tau protein for all subjects.

FIG. 5: A proposed flow chart to discriminate subjects in various disease groups by assaying multiple bio-markers.

FIG. 6: Detected values in the frame of plasma φ_Aβ1-42×φ_tau and plasma φ_α-syn for all subjects.

FIG. 7: Illustration of the polar angle θ to discriminate disease groups in the frame of plasma φ_Aβ1-42×φ_tau and plasma φ_α-syn for all subjects.

FIG. 8: Distribution of original concentrations, log(φ_α-syn) and log(φ_Aβ1-42×φ_tau) for variously disease groups.

FIG. 9: Distribution of modulated concentrations in the frame of φ′_α-syn and φ′_Aβ1-42×φ_tau for variously disease groups.

FIG. 10: Shifted polar angles θ′_s for NC group and DLB group in FIG. 9.

FIG. 11: Boundary guided with the ray along the polar angle 189.48° optimally discriminating NC and DLB subjects.

FIG. 12: Boundaries denoted with rays along specific polar angles to optimally discriminating subjects with NC (•), DLB (♦), PD (▴), AD (+), or FTD (▪).

SUMMARY OF THE INVENTION

The present invention relates to a method for constructing quadrants corresponding to different diseases in a frame of concentrations of multiple independent biomarkers, comprising:

• • (a) transferring original distributed concentrations of every independent biomarker to modified distributed concentrations, comprising:
    • calculating the mean value and the standard deviation of the original distributed concentrations for a given independent biomarker;
    • individually subtracting all the original distributed concentrations by the mean value for the given independent biomarker; and
    • individually dividing all the differences by the standard deviation to get the modified distributed concentrations for the given independent biomarker;
  • (b) positing the modified distributed concentrations in a frame of multiple independent biomarkers; and
  • (c) finding a boundary optimally separating neighboring quadrants corresponding to different diseases.

DETAILED DESCRIPTION OF THE INVENTION

It is proposed that DLB patients can be discriminated from NC and AD, as illustrated in FIG. 2. By firstly assaying plasma α-synuclein and then assaying plasma Aβ_1-42 and tau protein, DLB patients show higher φ_α-syn and lower φ_Aβ1-42×φ_tau, AD patients show higher φ_Aβ1-42×φ_tau, and NC subjects show lower φ_α-syn and lower φ_Aβ1-42×φ_tau. The illustration in FIG. 2 reveals the possibility to achieve high clinical sensitivity and specificity in discriminating patients with DLB or other neurodegenerative diseases such as AD, PD, and FTD by assaying multiple biomarkers.

Therefore, the present invention is to investigate a method for discriminating patients with different diseases by assaying multiple biomarkers.

The present invention provides a method for constructing quadrants corresponding to different diseases in a frame of concentrations of multiple independent biomarkers, comprising:

• • (a) transferring original distributed concentrations of every independent biomarker to modified distributed concentrations, comprising:
    • calculating the mean value and the standard deviation of the original distributed concentrations for a given independent biomarker;
    • individually subtracting all the original distributed concentrations by the mean value for the given independent biomarker; and
    • individually dividing all the differences by the standard deviation to get the modified distributed concentrations for the given independent biomarker;
  • (b) positing the modified distributed concentrations in a frame of multiple independent biomarkers; and
  • (c) finding a boundary optimally separating neighboring quadrants corresponding to different diseases.

In an embodiment, the original distributed concentrations are either originally detected concentrations or transferred concentrations of the originally detected concentrations via mathematic calculation. The mathematic calculation may be, but not limited to, logarithm, trigonometric function, sigmoid function, or any combinations thereof.

In an embodiment, the independent biomarker may be either a single kind of bio-molecule or a combination of several kinds of bio-molecules. The independent biomarker may exist in, but not limited to, blood, urine, cerebrospinal fluid, or saliva.

In an embodiment, the number of the multiple independent biomarkers is more than one.

In an embodiment, the method for finding a boundary optimally separating neighboring quadrants corresponding to different diseases in step (c) is ROC curve analysis.

In an embodiment, the independent biomarkers may be, but not limited to, plasma α-synuclein, β-amyloid, tau protein, and their derivatives to construct quadrants corresponding to healthy subjects and patients with neurodegenerative disease. The neurodegenerative disease may be, but not limited to, dementia with Lewy body, Parkinson's disease, Alzheimer's disease, or frontotemporal dementia.

EXAMPLES

The examples below are non-limiting and are merely representative of various aspects and features of the present invention. Therefore, the method of the present invention should not be limited to only for diagnosing neurodegenerative disease.

Example 1

Subjects

Subjects including normal controls (n=6), DLB patients (n=9), AD patients (n=6), PD patients (n=9), and FTD patients (n=6) were enrolled. The ages of all subjects were from 47 to 87 years. The demographic information of subjects was listed in Table 1. The subjects were divided into disease groups according to neuropsychological tests and clinical symptoms. Some subjects were examined with magnetic resonance imaging. There was no combination of these diseases for any one of subjects. The plasma biomarkers Aβ_1-42, tau protein, and α-synuclein were assayed for each subject by using immunomagnetic reduction (Yang, C. C., Yang, S. Y., Chieh, J. J., Horng, H. E., Hong, C. Y., Yang, H. C., Chen, K. H., Shih, B. Y., Chen, T. F., Chiu, M. J. (2011) Biofunctionalized magnetic nanoparticles for specifically detecting biomarkers of Alzheimer's disease in vitro. ACS Chem. Neurosci. 2, 500-505; Chiu, M. J., Yang, S. Y., Chen, T. F., Chieh, J. J., Huang, T. Z., Yip, P. K., Yang, H. C., Cheng, T. W., Chen, Y. F., Hua, M. S., and Horng, H. E. (2012) New assay for old markers-plasma beta amyloid of mild cognitive impairment and Alzheimer's disease. Curr. Alzheimer Res. 9, 1142-1147; Yang, S. Y., Chieh, J. J., Yang, C. C., Liao, S. H., Chen, H. H., Horng, H. E., Yang, H. C., Hong, C. Y., Chiu, M. J., Chen, T. F., Huang, K. W., and Wu, C. C. (2013) Clinic applications in assaying ultra-low-concentration bio-markers using HTS SQUID-based AC magnetosusceptometer. IEEE Trans. Appl. Supercond. 23, 1600604-1600607; Chiu, M. J., Chen, Y. F., Chen, T. F., Yang, S. Y., Yang, F. P. Gloria, Tseng, T. W., Chieh, J. J., Chen, J. C. Rare, Tzen, K. Y., Hua, M. S., and Horng, H. E. (2013) Plasma tau as a window to the brain-negative associations with brain volume and memory function in mild cognitive impairment and early alzheimer's disease. Human Barin Mapping, 35, 3132-3142; Tzen, K. Y., Yang, S. Y., Chen, T. F., Cheng, T. W., Horng, H. E., Wen, H. P., Huang, Y. Y., Shiue, C. Y., and Chiu, M. J. (2014) Plasma Aβ but not tau related to brain PiB retention in early Alzheimer's disease. ACS Neuro. Chem. 5, 830 (2014)).

TABLE 1
Demographic information of enrolled subjects.
	Group	Numbers (Male/Female)	Age (years)
	NC	6 (2/4)	53-73
	DLB	9 (6/3)	65-81
	AD	6 (3/3)	65-87
	PD	9 (5/4)	65-81
	FTD	6 (0/6)	47-73
	NC: normal control,
	DLB: dementia with Lewy Body
	AD: Alzheimer's disease,
	PD: Parkinson's disease,
	FTD: frontotemporal disease

Assay Results

The detected concentrations of Aβ_1-42, tau protein, and α-synuclein in plasma for all the subjects were shown in FIGS. 3A-3C. The detected concentrations of Aβ_1-42, tau protein, and α-synuclein in plasma were denoted as φ_Aβ1-42, φ_tau, φ_α-syn, respectively.

In FIG. 3A, almost all the detected concentrations of plasma Aβ_1-42 were below 20 pg/ml, except some data points with AD and PD. It was not easy to discriminate patients with various diseases according to the concentration of plasma Aβ_1-42. As to the detected concentration of plasma tau protein in FIG. 3B, NC showed relatively low concentrations of plasma tau protein, while AD and FTD showed relatively high concentrations. In case, it was not significant to differentiate DLB from PD, as well as AD from FTD. In FIG. 3C, PD could be obviously recognized due to the highest level for the concentrations of plasma α-synuclein as compared to other groups. DLB showed the second highest level for the concentrations of plasma α-synuclein. FTD was clearly recognized because of the definitely lowest level for the concentrations of plasma α-synuclein. However, the plasma α-synuclein for NC and AD distributed the same concentration range. It failed to discriminate NC and AD by using the concentration of plasma α-synuclein.

The results in FIGS. 3A-3C revealed the fact similar to that in FIGS. 1A and 1B that it failed to differentiate patients with various diseases by using single bio-marker. Multiple bio-markers were necessarily developed to achieve high discrimination among diseases.

Multiple Bio-Markers

It had been reported that the product in concentrations, denoted as φ_Aβ1-42×φ_tau, of plasma Aβ_1-42 and tau protein was superior to either of plasma Aβ_1-42 or plasma tau protein for diagnosing neurodegenerative disease, especially AD. The φ_Aβ1-42×φ_tau for all subjects were plotted in FIG. 4. The AD and FTD patients showed higher values for φ_Aβ1-42×φ_tau as compared to NC, DLB, and PD. With the results in FIG. 3C, FTD patients showed the lowest level for the concentrations of plasma α-synuclein. Thus, it was able to differentiate FTD from AD by additionally assaying plasma α-synuclein after finding φ_Aβ1-42×φ_tau.

For the groups with lower values of φ_Aβ1-42×φ_tau, i.e. NC, DLB, and PD, they showed different levels of φ_α-syn, as shown in FIG. 3C. PD patients showed the highest level of φ_α-syn, while NC subjects showed the lowest level of φ_α-syn. Hence, it was able to discriminate PD, DLB, and NC by additionally assaying plasma α-synuclein after finding φ_Aβ1-42×φ_tau. In summary, a flow chart for discriminating subjects of NC, DLB, AD, PD, and FTD by assaying plasma Aβ_1-42, tau protein, and α-synuclein was proposed and shown in FIG. 5.

2D Map for Disease Discrimination

According to FIG. 5, plasma φ_Aβ1-42×φ_tau and plasma φ_α-syn could be used as diagnostic parameters for discriminating NC, DLB, AD, PD, and FTD. It provided a motivation to construct the two-dimensional distribution in the frame of plasma φ_Aβ1-42×φ_tau and plasma φ_α-syn, as shown in FIG. 6. In FIG. 6, the data points for various groups were distributed in various region. Roughly speaking, standing at the center of the region occupied with the data points, NC was distributed in the lower-left region, FTD was distributed in the lower-right region, AD was roughly in the right region, PD was distributed in the upper region, and DLB was distributed in the upper-left region. It seemed possible to use a polar angle θ as a parameter to discriminate subjects into various disease groups, as illustrated in FIG. 7.

The polar angle θ was the angle span by a ray with respect to the horizontal ray. The horizontal ray was not the x axis in the frame of plasma φ_Aβ1-42×φ_tau and plasma φ_α-syn, but started from the center of the region occupied with the data points. It was necessary to find the starting point of the horizontal ray. The coordinates of the starting point of the horizontal ray could be defined as the averaged values of φ_Aβ1-42×φ_tau and φ_α-syn of all subjects. However, the detected values of φ_Aβ1-42×φ_tau or φ_α-syn distributed over several orders of magnitude. Points with higher values would be more weighted when calculating the average value of all points. This was the reason why FIG. 6 or FIG. 7 was not plotted in the linear-to-linear scale. Instead, FIG. 6 and FIG. 7 were plotted in the log-to-log scale.

The detected values of all points in FIG. 6 were transferred via logarithm. The log(φ_α-syn) versus log(φ_Aβ1-42×φ_tau) were plotted in FIG. 8. Hereafter, the values of log(φ_α-syn) and log(φ_Aβ1-42×φ_tau) were referred as to original concentrations. It was found that the values of all log(φ_α-syn)'s or all log(φ_Aβ1-42×φ_tau)'s were distributed at the same order of magnitude. For example, log(φ_α-syn)'s were within the range from −6 to 2, and log(φ_Aβ1-42×φ_tau)'s were within the range from 1.6 to 3.6. Now, each data point was equally weighted when calculating their averaged values of log(φ_α-syn)'s and log(φ_Aβ1-42×φ_tau)'s. In addition, the standard deviations of all log(φ_α-syn)'s and all log(φ_Aβ1-42×φ_tau)'s were calculated. Then, each log(φ_α-syn) and log(φ_Aβ1-42×φ_tau) were transferred via the following equations:

φ′_Aβ1-42×tau=[log(φ_Aβ1-42×φ_tau)−M_log(φ_Aβ1-42×φ_tau)]/SD_log(φ_Aβ1-42×φ_tau),   (Eq. 1)

φ′_α-syn=[log(φ_α-syn)−M_log(φ_α-syn)]/SD_log(φ_α-syn),   (Eq. 2)

Where M_log(φ_Aβ1-42×φ_tau) and SD_log(φ_Aβ1-42×φ_tau) denoted the averaged value and standard deviation of all log(φ_Aβ1-42×φ_tau)'s, M_log(φ_α-syn) and SD_log(φ_α-syn) denoted the averaged value and standard deviation of all log(φ_α-syn)'s. The φ′_Aβ1-42×tau and φ′_α-syn, which were referred as to modulated concentrations, were plotted in FIG. 9.

The origin in the frame of φ′_α-syn and φ′_Aβ1-42×tau was the starting point of the horizontal ray in FIG. 7. That was, the horizontal ray was the x axis of FIG. 9.

Polar Angle for Disease Discrimination

In FIG. 9, there existed a boundary which optimally separated subjects in two disease groups locating in neighboring regions. The boundary could be determined by suitably constructing a ray along a certain polar angle θ′ from the origin in the frame of φ′_α-syn and φ′_Aβ1-42×tau. An example to construct boundary optimally separating NC and DLB was given. Other boundaries for optimally separating two neighboring groups in FIG. 9 can be constructed by following the same processes.

Firstly, each point of NC group and DLB group was assigned with a polar angle θ′. For a given point i with the coordinate (φ′_{Aβ1-42×tau,i}, φ′_α-syn,i) in FIG. 9, the polar angle θ′_i was obtained via

θ′_i=tan⁻¹(φ′_α-syn,i/φ′_{Aβ1-42×tau,i})

Among the points in NC and DLB groups, there existed a minimum polar angle, referred as θ′_min. Every θ′_i was subtracted with θ′_min to get the shifted polar angle θ′_s,i, i.e.

θ′_s,i=θ′_i−θ′_min

The θ′_s's for points belonging to NC and DLB groups in FIG. 9 were plotted in FIG. 10. Through the analysis of receiver operating characteristic (ROC) curve, the cut-off value of θ′_s to optimally discriminate NC and DLB subjects was found to be 225.28°, as plotted with a dashed line in FIG. 10. The corresponding sensitivity and specificity were 0.833 and 1.000. By adding θ′_min to 225.28°, the polar angle optimally discriminating NC and DLB groups in FIG. 9 was 189.48°. The ray along the polar angle 189.48° was plotted with the solid line in FIG. 11.

The boundaries optimally separating neighboring groups in FIG. 9 could be found, as plotted with solid lines in FIG. 12.

Utilization of 2D Map for Disease Diagnosis

With the boundaries in FIG. 12, the two-dimensional map in the frame of φ′_α-syn and φ′_Aβ1-42×tau were divided into several quadrants. Each quadrant corresponded to one kind of diseases. Once the concentrations of plasma Aβ_1-42, tau protein, and α-synuclein of a subject had been detected, the modulated concentrations could be obtained via Eqs. (1) and (2). Then, the position of modulated concentrations with this subject could be found in FIG. 12. The subject could be accurately diagnosed.

One skilled in the art readily appreciates that the present invention is well adapted to carry out the objects and obtain the ends and advantages mentioned, as well as those inherent therein. The biomarkers and uses thereof are representative of preferred embodiments, are exemplary, and are not intended as limitations on the scope of the invention. Modifications therein and other uses will occur to those skilled in the art. These modifications are encompassed within the spirit of the invention and are defined by the scope of the claims.

It will be readily apparent to a person skilled in the art that varying substitutions and modifications may be made to the invention disclosed herein without departing from the scope and spirit of the invention.

All patents and publications mentioned in the specification are indicative of the levels of those of ordinary skill in the art to which the invention pertains. All patents and publications are herein incorporated by reference to the same extent as if each individual publication was specifically and individually indicated to be incorporated by reference.

The invention illustratively described herein suitably may be practiced in the absence of any element or elements, limitation or limitations, which are not specifically disclosed herein. The terms and expressions which have been employed are used as terms of description and not of limitation, and there is no intention that in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, but it is recognized that various modifications are possible within the scope of the invention claimed. Thus, it should be understood that although the present invention has been specifically disclosed by preferred embodiments and optional features, modification and variation of the concepts herein disclosed may be resorted to by those skilled in the art, and that such modifications and variations are considered to be within the scope of this invention as defined by the appended claims.

Claims

1. A method for constructing quadrants corresponding to different diseases in a frame of concentrations of multiple independent biomarkers, comprising: (a) transferring original distributed concentrations of every independent biomarker to modified distributed concentrations, comprising: calculating the mean value and the standard deviation of the original distributed concentrations for a given independent biomarker;individually subtracting all the original distributed concentrations by the mean value for the given independent biomarker; andindividually dividing all the differences by the standard deviation to get the modified distributed concentrations for the given independent biomarker;(b) positing the modified distributed concentrations in a frame of multiple independent biomarkers; and(c) finding a boundary optimally separating neighboring quadrants corresponding to different diseases.

2. The method of claim 1, wherein the original distributed concentrations are originally detected concentrations or transferred concentrations of the originally detected concentrations via mathematic calculation.

3. The method of claim 2, wherein the mathematic calculation is logarithm, trigonometric function, sigmoid function, or any combinations thereof.

4. The method of claim 1, wherein the independent biomarkers are selected from the group consisting of a single kind of bio-molecule and a combination of several kinds of bio-molecules.

5. The method of claim 1, wherein the independent biomarkers exist in blood, urine, cerebrospinal fluid, or saliva.

6. The method of claim 1, wherein the number of the multiple independent biomarkers is more than one.

7. The method of claim 1, wherein the method for finding a boundary optimally separating neighboring quadrants corresponding to different diseases in step (c) is ROC curve analysis.

8. The method of claim 1, wherein the independent biomarkers are plasma α-synuclein, β-amyloid, tau protein, and their derivatives to construct quadrants corresponding to healthy subjects and patients with neurodegenerative disease.

9. The method of claim 8, wherein the neurodegenerative disease is dementia with Lewy body, Parkinson's disease, Alzheimer's disease, or frontotemporal dementia.