You are on page 1of 4

Development, Evaluation and ID-GCMS Comparison of a Highly Sensitive

Automated Immunoassay for the Determination of Estradiol on the Abbott


ARCHITECT Analyzer
G. Williams1, F. Hayes2, S. Frost1, J.Ramp1, D. Pacenti1, W. Barnes1, R. Doss1, W. Crowley2, D. C. Lehotay3, S. George3,
J. Eichhorst3, C. Ramsay1, P. Sluss2
Abbott, Diagnostics Division, Abbott Park, IL 60064, U.S.A.; 2Reproductive Endocrine Unit, Massachusetts General Hospital, Boston, MA., 02114;
1

University of Saskatchewan College of Medicine, Provincial Laboratory SK S4S 5W6


3

XIX International Congress of Clinical Chemistry, IFCC/AACC 2005 Annual Meeting, July 24 – 28, 2005 – Orlando, Florida

Introduction
Measurement of estradiol (E2) plays a 1st Generation ARCHITECT E2 Assay
critical role in the clinical management of
infertility. During in vitro fertilization (IVF), E2 ������
levels may reach ≥3,000 pg/mL. There is
�����������������������

also a need to be able to measure low ������


levels of E2 (15 pg – 75 pg/mL); in pubertal
������������

girls, postmenopausal women and in men. �����


The challenge for all currently available
direct methods for measuring E2 in serum �����
is to provide accuracy (by insuring
calibration to ID-GCMS) and excellent �����

precision across a wide dynamic range.


�����
The 1st generation ARCHITECT E2
���
assay was developed for IVF testing
���� ����� ������ ������ ������ ������ ������ ������
where a wide dynamic range is needed to
�����������������������
measure relatively high [>1000 pg/mL] �������
concentrations. Similar to most direct
automated methods, this assay was too
This assay is no longer commercially available.
insensitive for other clinical applications
and was inaccurate at low concentrations
[<100 pg/mL].

Results
Using the same high-affinity monoclonal 2nd Generation ARCHITECT E2 Assay
anti-E2 and modified competitive assay
format, we developed a 2nd generation
����
������������������������������

ARCHITECT E2 assay. Design modifi-


cations, format, and tracer concentration
��������������������
had a major impact on accuracy. This ����
�����������������������


improvement is illustrated below. � ���������
���

���

���

���


� ��� ��� ��� ��� ����
�������
Results
The 2nd Generation E2 assay showed improved concordance with clinical expectations when used to measure E2 in specimens containing
low levels of E2. Figure 3 below shows the distribution of E2 measurements in post-menopausal women and men.

�� ��
������������������������������������������������������������� ��������������������������������������������
�� ������� ������
�� ������� ������
�����������
�����������
�� ����� ������
� ����� ������
���� ������ ���� ������
���� ������ ���� ������
� ���� ������

�����
���� ������
�����

� ������ ������� ������ �������



����������� ��������� ������ ����������� ��������� ������

����� ���� ������������ � ����� ��� �������������
������� ������ ���������������� ������ ������ ����������������

� �
�� ��
�������������������������������������������������������������
���������������������������������������������
�� ������� ������
����� ������ ��
�����������
�� ����������� ����� ������
����� ������ �� ���� ������
�����

�����
���� ������ ���� ������
�� ���� ������ ���� ������
���� ������ �� ������ ������
������ �������
�� ����������� ��������� ������
������ ��� ������������
����������� ��������� ������ �� ������ ������ ����������������
�� ������ ���� ���
������ ������ ����������������
� �

� �
� �� �� �� �� ��� ��� ��� ��� � �� �� �� �� ��� ���
����������������������� �����������������������

The 2nd Generation E2 assay is accurate based upon standardization to ID-GCMS (Clin.Chem.2002;48:364-366). Table 2 shows the
concordance between assay and “gold standard” measurement of E2 in individual clinical specimens spanning the dynamic range of
the assay.
Abbott ARCHITECT Estradiol vs. ID-GCMS
Range N Min GCMS (pg/mL) Max GCMS (pg/mL) r
10 to 1,000 pg/mL 315 13.3 945.2 0.99
10 to 5,000 pg/mL 393 13.3 4269.2 0.99

Slope per Reagent Lot


Regression Method N Pooled Intercept Pooled Slope Lot 1 Lot 2 Lot 3
10 – 1,000 pg/mL
Least Squares 105 0 1.03 0.99 1.03 1.08
Passing-Bablok 105 2 1.02 0.98 1.02 1.06
10 – 5,000 pg/mL
Least Squares 131 -31 1.09 1.08 1.08 1.12
Passing-Bablok 131 0 1.04 1 1.03 1.07
The 2nd Generation E2 assay had improved precision and sensitivity. Three control levels were used according to NCCLS protocol EP5-
A.18 to assess precision of the assay. Table 3 shows these results from three lots of reagents, tested in replicates of two at twice daily for
20 days on two instruments. Figure 5 shows functional sensitivity [CV<20%] which was <15 pg/mL.

Within Run Total


Control Reagent Lot Instrument N Mean (pg/mL) SD (pg/mL) %CV SD (pg/mL) %CV
Low 1 1 80 48 2.6 5.5 3.3 6.7
Low 1 2 80 47 2.4 5.0 2.8 6.0
Low 2 1 80 45 2.9 6.4 3.3 7.4
Low 2 2 80 46 2.1 4.6 2.6 5.6
Low 3 1 80 45 2.7 6.0 3.0 6.6
Low 3 2 80 47 2.3 4.8 3.3 7.1
Medium 1 1 80 188 3.5 1.9 4.0 2.1
Medium 1 2 80 188 3.6 1.9 4.3 2.3
Medium 2 1 80 188 4.2 2.3 4.6 2.4
Medium 2 2 80 190 3.3 1.7 4.3 2.3
Medium 3 1 80 183 3.7 2.0 4.5 2.5
Medium 3 2 80 192 3.0 1.5 4.3 2.3
High 1 1 80 596 10.5 1.8 10.5 1.8
High 1 2 80 593 8.5 1.4 10.6 1.8
High 2 1 80 606 9.8 1.6 11.5 1.9
High 2 2 80 604 10.3 1.7 10.9 1.8
High 3 1 80 590 10.9 1.8 13.7 2.3
High 3 2 80 591 10.7 1.8 15.2 2.6

��

������������������ ��� ������������������������


��

��
������������

��

��

��


� �� �� �� ��
�����������������
Results
Reference ranges were obtained by testing specimens [101 males, 72 postmenopausal females and 36 normal menstruating females]
from carefully characterized volunteers in a prospective clinical study. Variations in cycle length were normalized based on Day 0 = day LH
peak. Results are summarized in Table 4. Figure 6 shows daily reference ranges for ovulatory menstrual cycles.

Estradiol Value (pg/mL)


N Median Central 95% Range
Normal menstruating females
Follicular phase
385 54 21 – 251
(days –15 to –2)
Mid-cycle phase
105 196 38 – 649
(days –1 to +1)
Luteal phase
466 99 21 – 312
(days +2 to +15)

Post-menopausal females not on HRT 50 <10 <10 – 28

Post-menopausal females on HRT 22 28 <10 – 144

Males 101 23 11 – 44

ARCHITECT Estradiol Profile During


the Normal Menstrual Cycle

Conclusion
We describe a new, highly-sensitive, direct E2 assay on the ARCHITECT. The assay has
an auto-dilution range up to 5,000 pg/mL. Multiple lot precision was ≤7.4% (total CV) at the
45 pg/mL level, and ≤2.6% (total CV) for concentrations at the 190 pg/mL and 600 pg/mL
levels. The estimated functional sensitivity (CV≤20%), using human panels ranging from
0 to 70 pg/mL, was ≤14 pg/mL. The assay is standardized by ID-GCMS. This ARCHITECT
i2000 2nd generation E2 assay demonstrates significantly improved accuracy and
reproducibility across a clinically important range for improved diagnostic utility.

98-1118/R1 Jun 2005 Printed in U.S.A. 0526035_O

You might also like