i am the director of the northwest lipid metabolism and diabetes … · 2019-05-29 · eas 2019 i...

31
EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories (NWRL) at the University of Washington in Seattle WA, USA NWRL/UW has received grant and research funding from: National Institutes of Health Amgen Inc Ionic Pharmaceuticals Kaiser Permanente I am consultant to: Denka Seiken, Japan Roche Diagnostics, Germany Medtest DX Inc, USA

Upload: others

Post on 27-May-2020

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

EAS 2019

I am the Director of the Northwest Lipid Metabolism and

Diabetes Research Laboratories (NWRL) at the University of

Washington in Seattle WA, USA

NWRL/UW has received grant and research funding from:

• National Institutes of Health

• Amgen Inc

• Ionic Pharmaceuticals

• Kaiser Permanente

I am consultant to:

• Denka Seiken, Japan

• Roche Diagnostics, Germany

• Medtest DX Inc, USA

Page 2: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

Santica M Marcovina PhD ScD FAHAResearch Professor of Medicine, UW

Director, Northwest Lipid Metabolism and Diabetes Research Laboratories

EAS 2019

Page 3: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

Aim: To obtain values in patient samples that are accurate and

comparable independently of the analytical method used for the

measurement. Accuracy is obtained if calibrator values are

traceable to a high order primary reference material directly or via

a suitable secondary reference material and the methods are

certified to guarantee that accuracy of the calibrator results in

accurate values in samples.

Aim: To obtain values in patient samples that are comparable

even though not necessarily accurate.

EAS 2019

Page 4: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

• Antibody specificity

• Immunoreactivity of the antibody per particle should be the

same for the assay calibrator and for the samples

• An accuracy-based target value should be assigned to the

assay calibrators

EAS 2019

Page 5: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

274

399

649

Kringle 4 Type

V PD

Kringle 4

187

kDa

3-1021

Type 2 repeats

10

20

40

3

30 524

SM Marcovina, University of Washington EAS 2019

Page 6: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

EAS 2019

Page 7: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

• For Lp(a) the “signal” does not reflect the number of particles

• In samples with apo(a) sizes smaller than the apo(a) size of the

assay calibrator, Lp(a) levels will be

• In samples with apo(a) sized larger than the apo(a) size of the

assay calibrator, Lp(a) levels will be

UNDERESTIMATED

OVERESTIMATED

EAS 2019

Page 8: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

• To circumvent this problem, we produced a variety of monoclonal

antibodies against Lp(a) and we were able to obtain and characterize a

high-affinity monoclonal antibody directed to an epitome only present in

the KIV Type 9 region of apo(a)

• Using this unique monoclonal antibody (a-40), we developed and fully

validated an ELISA method that can measure Lp(a) protein on an equimolar

basis

• We demonstrated that Lp(a) contains one molecule of apo(a) per Lp(a)

particle and therefore, the Lp(a) protein values expressed in nmol/L reflect

the number of circulating Lp(a) particles.

EAS 2019

Page 9: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

EAS 2019

Calibrated with the same sample

containing apo(s) with 21 Kingle 4

Page 10: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

14161820222426283032-50

-25

0

25

50

Kringle 4 number

% Bias

y = 4.05x - 84.7

r = 0.967

EAS 2019

Page 11: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

• In the mid 1970’s, John Albers developed the first, high

sensitive RIA to measure Lp(a)

• Lp(a) was isolated from an individual with high Lp(a) levels

and measurements of the lipid and the protein components of

Lp(a) were performed

• This Lp(a) preparation was used as the RIA primary calibrator

with the assigned value being the sum of lipid and protein

values

• Results of Lp(a) were therefore expressed in mg/dL of TOTAL

MASS of Lp(a), not knowing that the Lp(a) mass is highly

variable within and between individuals

EAS 2019

Page 12: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

1997: IFCC Working Group

Members:

A. Steinmetz (Chair) F. Dati K. Berg

R. Couderc G. Kostner N. Rifai

I. Sakurabayashi J. Tate

Phase 1: Assessment of analytical performance of 40 test systems by testing serum

samples for precision, linearity, and parallelism.

Results:

❖ A significant number of assays were not optimized and failed to meet the criteria for

precision, linearity, and parallelism.

❖ The among method CV of all systems on reference samples ranged from 22% to 60%.

❖ The among method CV of optimized systems ranged from 16% to 35%

Clin Chem 1998; 44:1629-40

International Federation

of Clinical Chemistry

EAS 2019

Page 13: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

1998: Phase 2

Aims:

❖ Selection of a secondary reference material for Lp(a)

❖ Four proposed materials were compared in 27 optimized analytical systems

❖ Results of precision and linearity were comparable

❖ Among-method CV on each of the four materials ranged from 11% to 22%

Based on the overall results, the IFCC Working Group decided to select the material 2B as a

proposed reference material (PRM-2B) for further evaluation and for final target value

assignment. Decision was made to express Lp(a) values in nmol/L.

Clin Chem Lab Med 1999; 37:949-958

International Federation

of Clinical Chemistry

EAS 2019

Page 14: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

Principal Investigator:

❖ Santica M Marcovina,

University of Washington

Co-Investigators:

❖ John J Albers,

University of Washington

❖ Angelo Scanu,

University of Chicago

❖ Celina Edelstein,

University of Chicago

National Institutes of Health

National Heart, Lung, and Blood Institute

EAS 2019

Page 15: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

Specific Aims:

• Evaluation of Lp(a) assays with particular emphasis on

antibody specificity and sensitivity to apo(a) size

polymorphism

• Feasibility of standardization of Lp(a) assays

• Preparation and characterization of primary reference

material

• Assignment of target value to the IFCC Reference Material

National Institutes of Health

National Heart, Lung, and Blood Institute

EAS 2019

Page 16: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

• Study performed in collaboration with the IFCC Working

Group on Lp(a) standardization

• NWRL ELISA served as the reference method

• 21 commercially available methods were evaluated by using

30 fresh-frozen samples spanning a large range of Lp(a)

levels and apo(a) isoforms

• The reference material, PRM-2B, with an assigned value of

107 nmol/L, was used as a common assay calibrator

EAS 2019

Page 17: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

• Using PRM-2B to calibrate all the systems, the among-method CVs for each

of the 30 samples ranged from 6% to 31%

• Even though the CVs were lower than those obtained using individual

calibration (CV = 16% to 35%), no harmonization in Lp(a) values measured

by different methods was achieved

• The impact of apo(a) isoform size on Lp(a) concentrations varied among the

different methods as a function of the apo(a) size of the assay calibrators

• Other factors, like difference in instrumentation, also contributed to the

lack of comparability of results

• In 2003, the WHO Expert Committee on Biological Standardization accepted

PRM-2B, with an assigned value of 107 nmol/L, as the “First WHO/IFCC

International Reference Reagent for Lipoprotein(a) Immunoassay”

EAS 2019

Page 18: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

10 15 20 25 30 35-100

-50

0

50

100

Major Kringle 4 Repeat

Bias (%)

EAS 2019

Page 19: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

10 15 20 25 30 35-100

-50

0

50

100

Major Kringle 4 Repeat

Bias (%)

EAS 2019

Page 20: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

10 15 20 25 30 35-50

-25

0

25

50

Major Kringle 4 Repeat

Bias (%)

EAS 2019

Page 21: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

• Polyclonal antibody-based. Antibodies strongly reacting with

apo(a) KIV Type 2 repeats.

• Antibodies are coated on the surface of latex particles

(diameter 120 nm)

• Lp(a) particles in plasma are then bound to the antibodies on

the latex particles to form immunocomplexes

• Five independent calibrators with Lp(a) levels from low to high

and apo(a) size from large to small

EAS 2019

Page 22: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

Very

Small

Medium

Small

Medium

Large

Very

Large

Lp(a) nmol/L

Signal

EAS 2019

Page 23: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

Values expressed in mg/dL of total Lp(a) mass

➔ Single calibrator: no traceability to a common

reference material

• Effect of apo(a) size variation is inevitable

Values traceable to the WHO/IFCC Reference Material

and expressed in nmol/L of Lp(a) protein

➔ 5-point calibrators

• Effect of apo(a) size variation is minimized

EAS 2019

Page 24: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

apo(a) Kringle number accounts for 61.1% of the bias variation

ITAITA-DENKA

apo(a) Kringle number accounts for 2.5% of the bias variation

EAS 2019

Page 25: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

Denka Method – Instrument 1

EAS 2019

Denka Method – Instrument 2

apo(a) Kringle number accounts for 1.0% of the bias variation apo(a) Kringle number accounts for 0.7% of the bias variation

Page 26: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

Step 1 – Assignment of target values to assay calibrators

• Target values traceable to the WHO-IFCC Reference Material

Step 2 – Validation of the accuracy of value transfer

• Six fresh-frozen samples from individual donors were

prepared at the NWRL and were selected to have a suitable

range of Lp(a) levels and apo(a) isoforms.

EAS 2019

Page 27: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

Step 3

• 80 fresh-frozen samples are analyzed, collected from

apparently healthy donors and selected to represent a large

range of Lp(a) levels and apo(a) isoforms

• Detailed certification criteria have been established including

analytical CV, correlation and absolute bias limits between

obtained and assigned values, and the correlation between

the % bias and apo(a) isoform size

EAS 2019

Page 28: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

System 1 System 2

System 3 System 4

EAS 2019

Page 29: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

• To guarantee that a new lot of reagent and calibrator results in

minimization of the effect of apo(a) size variability on Lp(a) values obtained

by different analytical platforms, requires rigorous work for the distributors

of the Denka assay. Clinical chemistry laboratories using the assay should

ask for demonstration that assay optimization has been performed on their

specific instrument.

• The limitations of the assay should be understood by clinicians and by

clinical chemistry laboratories.

• Claims that the assay is accurate because Denka antibodies do not

recognize the KIV Type 2 repeats of apo(a) or that the values produced by

the assay are not affected by apo(a) size polymorphism are simply false

claims.

• Reviewers of scientific publications should reject false claims about this

assay and request specific demonstration that the impact of apo(a)

variability on their data has been minimized on the instrument used for

performing the analyses.

EAS 2019

Page 30: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

• The first small step toward standardization at this point in time is the expression of

Lp(a) values in nmol/L of Lp(a) protein considering that the assays measure apo(a)

and not the total Lp(a) mass.

• Even though the nmol/L protein determined using antibodies, either polyclonal or

monoclonal, directed against the variably expressed KIV Type 2 repeats do not

accurately reflect the number of circulating particles, a common unit for expressing

Lp(a) levels is not only scientifically correct but it’s advisable to avoid the use of

arbitrary, incorrect, and confusing conversion factors to transform values obtained

by different methods.

• The WHO/IFCC Reference Materials is available free of charge from our laboratory

upon request from any manufacturer who wants to change their calibrator units in

nmol/L.

• Manufacturers and clinicians should be aware that traceability of the calibrator to

the reference material does not indicate accuracy of values in patient samples.

EAS 2019

Page 31: I am the Director of the Northwest Lipid Metabolism and Diabetes … · 2019-05-29 · EAS 2019 I am the Director of the Northwest Lipid Metabolism and Diabetes Research Laboratories

• Little has changed in the arena of Lp(a) measurements since the first attempt by the

IFCC WG to standardize Lp(a) methods by producing the WHO/IFCC First Reference

Material over 20 years ago.

• The problems of Lp(a) measurement have been clearly defined and some methods

able to minimize the effect of apo(a) size variation have been optimized.

• However, we are far away from standardization, or even harmonization, of Lp(a)

values obtained by different methods because the main problem for Lp(a) is not in

the assay calibrator but in the variable immunoreactivity of the assay antibodies

which cannot be eliminated.

• No new reference material is needed at present because no assay will benefit from

its introduction.

• Effort needs to be directed to the development of new approaches to measure Lp(a)

that are antibody-independent, precise, high throughput, and are available to

clinical chemistry laboratories.

• In conjunction to this type of automated instruments, a validated reference material

and an accuracy-based reference method will be necessary.

EAS 2019