close

Вход

Забыли?

вход по аккаунту

?

A multicenter investigation with interphase fluorescence in situ hybridization using X- and Y-chromosome probes

код для вставкиСкачать
American Journal of Medical Genetics 76:318–326 (1998)
A Multicenter Investigation With Interphase
Fluorescence In Situ Hybridization Using X- and
Y-Chromosome Probes
Gordon Dewald,1* Richard Stallard,2 A. Al Saadi,3 Susan Arnold,4 Patricia I. Bader,5
Ruthann Blough,6 Kathy Chen,7 B. Rafael Elejalde,8 Catherine J. Harris,9 Rodney R. Higgins,10
Gerald A. Hoeltge,11 Wei-Tong Hsu,12 Virginia Kubic,13 D. James McCorquodale,14 Mark A. Micale,15
J.W. Moore,16 Rosalie M. Phillips,17 Susan Scheib-Wixted,18 Stuart Schwartz,19 Steven Siembieda,20
Kathy Strole,21 Peter VanTuinen,22 Gail H. Vance,23 Ann Wiktor,24 Laura Wise,25 Jar-Fee Yung,26
Julie Zenger-Hain,27 and Alan Zinsmeister1
1
Mayo Clinic, Rochester, Minnesota
Great Lakes Regional Genetics Group, University of Wisconsin, Madison, Wisconsin
3
William Beaumont Hospital, Royal Oak, Michigan
4
Children’s Hospital Medical Center, Akron, Ohio
5
Parkview Memorial Hospital, Fort Wayne, Indiana
6
Children’s Hospital Medical Center, Cincinnati, Ohio
7
Lutheran General Hospital, Park Ridge, Illinois
8
Medical Genetics Institute, S.C., Milwaukee, Wisconsin
9
Cook County Hospital, Chicago, Illinois
10
Allina Health System, Minneapolis, Minnesota
11
The Cleveland Clinic Foundation, Cleveland, Ohio
12
Rush-Presbyterian-St. Luke’s Medical Center, Chicago, Illinois
13
Hennepin County Medical Center, Minneapolis, Minnesota
14
Michael Reese Hospital, Chicago, Illinois
15
Medical College of Ohio, Toledo, Ohio
16
Children’s Hospital, Columbus, Ohio
17
University of Minnesota Hospitals and Clinics, Minneapolis, Minnesota
18
University of Wisconsin, Madison, Wisconsin
19
University Hospitals, Cleveland, Ohio
20
St. Paul Ramsey Medical Center, St. Paul, Minnesota
21
Butterworth Hospital, Grand Rapids, Michigan
22
Medical College of Wisconsin, Milwaukee, Wisconsin
23
Indiana University, Indianapolis, Indiana
24
Henry Ford Hospital, Detroit, Michigan
25
Advanced Institute of Fertility, Milwaukee, Wisconsin
26
Mercy Hospital and Medical Center, Chicago, Illinois
27
Oakwood Hospital Laboratory, Dearborn, Michigan
2
Twenty-six laboratories used X and Y chromosome probes and the same procedures to
process and examine 15,600 metaphases and
49,400 interphases from Phaseolus vulgarisleucoagglutinin (PHA)-stimulated lymphocytes. In Part I, each laboratory scored 50
metaphases and 200 interphases from a normal male and a normal female from its own
practice. In Part II, each laboratory scored
Contract grant sponsor: Maternal and Child Health Bureau;
Contract grant number: MCJ-551004-04.
*Correspondence to: Gordon Dewald, Ph.D., Cytogenetics Laboratory, Mayo Clinic, 200 First Street SW, Rochester, MN 55905.
Received 19 August 1997; Accepted 24 December 1997
© 1998 Wiley-Liss, Inc.
50 metaphases and 200 interphases on slides
prepared by a central laboratory from a normal male and a normal female and three
mixtures of cells from the male and female.
In Part III, each laboratory scored 50 metaphases (in samples of 5, 10, 15, and 20) and
100 interphases (in samples of 5, 10, 15, 20,
and 50) on new, coded slides of the same
specimens used in Part II. Metaphases from
male specimens were scored as 98–99% XY
with no XX cells, and 97–98% of interphases
were scored as XY with 0.04% XX cells. Metaphases from female specimens were scored
as 96–97% XX with 0.03% XY cells, and 94–
96% of interphases were scored as XX with
0.05% XY cells. Considering the data as a
Multicenter Study With Interphase FISH
model for any probe used with fluorescence
in situ hybridization (FISH), a statistical approach assessing the impact of analytical
sensitivity on the numbers of observations
required to assay for potential mosaicisms
and chimerisms is discussed. The workload
associated with processing slides and scoring 50 metaphases and 200 interphases using FISH averaged 27.1 and 28.6 minutes,
respectively. This study indicates that multiple laboratories can test/develop guidelines for the rapid, efficacious, and costeffective integration of FISH into clinical
service. Am. J. Med. Genet. 76:318–326, 1998.
© 1998 Wiley-Liss, Inc.
KEY WORDS: mosaicism; X- and Y-chromosome probes; interphase fluorescence in situ hybridization; quality assurance; analytical sensitivity; workload;
chimerism; clone
INTRODUCTION
Fluorescence in situ hybridization (FISH) has become an integral part of clinical practice in cytogenetics. DNA probes are available as chromosome-specific
paints (CSP), chromosome-specific alpha satellites (alphoids), and certain locus-specific sites (LS). FISH is
used to examine cells in metaphase and interphase and
can be applied to almost any type of tissue [Trask,
1991]. However, the reproducibility, normal range, and
accuracy of FISH in clinical practice has been established for only a few probes [Jenkins et al., 1992; Dewald et al., 1993a,b; Lu et al., 1994; Schad and Dewald,
1995].
Because the application of FISH in clinical practice
was proceeding rapidly, the Standing Committee for
Cytogenetic Nomenclature developed formal nomenclature for the technology [ISCN, 1995]. The American
College of Medical Genetics (ACMG) produced guidelines for validation of FISH in clinical practice [American College of Medical Genetics, 1996]. Also, the
ACMG and the College of American Pathologists conducted pilot studies in preparation for national proficiency testing with FISH [Dewald et al., 1997]. Although many publications have demonstrated the accuracy of FISH, the Food and Drug Administration has
cleared only a few DNA probes for clinical practice
[Koska, 1997].
In 1995, 23 laboratories in the Great Lakes Regional
Genetics Group (GLaRGG) collaborated to use FISH to
evaluate a probe for small nuclear ribonucleoprotein
polypeptide N (SNRPN) in metaphases from normal
individuals and from patients with the Prader-Willi
syndrome and a visible deletion in proximal 15q [Dewald et al., 1996]. In that study, multiple centers produced equivalent data from their own slides and from
externally supplied slides. The centers also worked on
the same samples to assess workload and to test guidelines for metaphase FISH. In the present investigation,
319
26 cytogenetic laboratories in GLaRGG organized to
collect data for validating two probes for interphase
FISH, to measure the effort (workload) associated with
interphase FISH, and to test the efficacy of interphase
FISH in clinical service.
MATERIALS AND METHODS
Participants
Fifty-seven institutions providing clinical cytogenetic services within the states covered by GLaRGG
(IL, IN, MI, MN, OH, and WI) were invited to join a
project involving FISH. Of 39 respondents, 26 volunteered to follow a protocol and to dedicate personnel to
the project. Twenty-one of the volunteers participated
in an earlier study with FISH and a probe for the SNRPN locus [Dewald et al., 1996]. Although they were
self-selected, participants comprised a reasonable
sample of cytogenetic laboratories in the United States.
Equipment, Experience, and Organization
By using a questionnaire, participants reported the
filters, wattages, manufacturers, models, and lenses of
microscopes used for FISH. Participants also disclosed
the number of times chromosome-specific paints, alphoids, and locus-specific probes had been used in 1995
for clinical evaluations of amniotic fluid, bone marrow,
and peripheral blood. Each application of a probe to a
sample counted as a separate utilization.
Direct-labeled, differently colored probes were chosen for in situ hybridization because of equal utility
with condensed chromosomes (metaphases) or nuclei
(interphases) and because of their large, bright signals.
The probe for the X chromosome (DXZ1) is specific for
the pericentromeric segment, whereas the probe for the
Y chromosome (DYZ1) is specific for the heterochromatic segment of Yq12. Vysis, Inc. (Downers Grove, IL)
provided participants with probes, reagents, and detailed instructions for preparing and storing slides, denaturing DNA, hybridizing in situ, mixing reagents,
and scoring fluorescent signals (package insert, Spectrum CEP-X/Y dual-color probe kit; Vysis, Inc.). Volunteers agreed to follow the instructions as a condition of
participation. Four modifications were made to the
manufacturer’s instructions during Part I of the project: (1) Specimens were hybridized in moist chambers
overnight; (2) during hybridizations, coverslips were
sealed to slides with rubber cement; (3) samples of interphases were 100 and 200; and (4) some participants
diluted the propidium iodide counterstain.
A general plan for the course of the three-part project
with stimulated lymphocytes was sent to participants.
Data collection occurred between March 5 and April 15,
1996. In Part I, participants scored 50 metaphases and
200 interphases from slides they prepared from a chromosomally normal male (specimen 1) and female
(specimen 2) originating in their local practices.
In Part II, a central laboratory prepared slides [Spurbeck et al., 1996] from a chromosomally normal male
(specimen 3) and female (specimen 4). The central laboratory also prepared slides from artificial chimeras,
which were made by mixing suspensions of the cultured cells from specimens 3 and 4 in different proportions. The artificial chimeras were designed to contain
320
Dewald et al.
about 45%, 20%, and 8% XY cells (specimens 5–7, respectively). The five specimens were coded and sent in
duplicate to each participant. They were described only
as a male, a female, and three mixtures. Participants
processed and scored 50 metaphases and 200 interphases per specimen.
In Part III, the central laboratory provided new
slides (specimens 8–12) to participants. These were
prepared at the same time and from the same specimens used in Part II (i.e., repeat specimens of the normal male and female and the three artificial chimeras),
but they were coded and blinded for participants.
These specimens were described only as having unknown frequencies of XX and XY cells. Fifty metaphases (in samples of 5, 10, 15, and 20) and 100 interphases (in samples of 5, 10, 15, 20, and 50) were scored
per specimen in Part III.
Instructions and forms were sent to participants at
appropriate times to help assure uniform collecting and
reporting of data. Participants supplying incomplete or
excess numbers of observations were asked to reexamine stored slides to supply the designated observations.
Results were summarized and distributed so that participants could review their own results and compare
them anonymously with those of other participants.
Calculations
Metaphases or interphases with at least one fluorescent signal were scored serially until a predetermined
number of observations had been made. The sum of
observations with XY signals divided by the total number of cells scored by all participants for specimens 1, 3,
and 9 provided analytical sensitivities of the CEP-X/Y
probes in metaphases and interphases from males. The
corresponding analytical sensitivities for metaphases
and interphases from females, likewise, were computed
from the sums of cells with XX signals in specimens 2,
4, and 10.
Mean percentages (total number of observations/
total number of cells scored by all participants) and
ranges (actual observations of individual participants
expressed as percentages) were used to summarize
data collected by the 26 participants from the normal
specimens and the artificial chimeras. To assess the
consistency of the data, normal (i.e., target) ranges
were constructed by omitting the high and low frequencies of XY, XX, and the various non-XY/non-XX signals
in samples from the chromosomally normal male
(specimen 3) and female (specimen 4) studied by all
participants. Similarly, normal ranges were constructed for the artificial chimeras by omitting the high
and low frequencies of XY and XX signals in samples
from specimens 5–7. The resulting normal ranges approximated 92% of the distributions of the individual
values reported by the 26 participants.
The following formula was used to estimate the number of observations (N) required to provide sufficient
power (95% or 99%) to detect mosaicisms or chimerisms (Table VII). The formula is based on a one-sided,
hypothesis-testing approach for a binomial proportion.
N=
F
Za=po~1 − po! + Zb=pa~1 − pa!
po − pa
G
2
.
In this formula, Za ≡ 1.645 for a (one-sided) alpha-level
test yielding a Type I error rate of 5% (i.e., rejecting the
null hypothesis, Ho, when Ho is true); Zb ≡ 1.645 for
95% power, or Zb ≡ 2.330 for 99% power; po 4 the
analytical sensitivity of the probe, i.e., the proportion of
signals (e.g., XY or XX) obtained from normal specimens; and pa 4 the frequency of the more common cell
line, i.e., the actual or potential proportion of signals
(e.g., XY or XX) present in the test specimen.
The formula provided above was based on the binomial distribution, a distribution that was also used to
compute the probability of observing one, two, three, up
to N cells of a potential second cell line in samples of
various sizes (Table VIII). These probabilities were calculated by assuming that the proportion of cells in the
second cell line was one minus the analytical sensitivity (i.e., 1 − po).
Workload Recording
Forms for recording the workload associated with
FISH followed the manufacturer’s instructions for using the probes. Each participant measured the time
invested in 57 specified steps. To provide preliminary
experience with processing, observing, and documenting (including capturing two images digitally or photographically) before the actual timings, participants
treated specimens in numerical sequence. Participants
provided timings in minutes and seconds only for
single slides from specimens 6 and 7, but they reported
the processing of any backup slide required to score 50
metaphases and 200 interphases.
RESULTS
Equipment
Participants employed various microscopes: 12 used
Zeiss (Thornwood, NY), seven used Olympus (Tokyo,
Japan), four used Leitz (Wetzlar, Germany) or Leica
(Deerfield, IL), and three Nikon (Tokyo, Japan). Onehundred-watt bulbs provided illumination for 23 participants, and 50-watt bulbs provided illumination for
three participants. Twenty-four participants viewed
signals with triple band-pass filters, whereas two observed with dual band-pass filters. Participants made
observations through various combinations of objective
lenses: Five used only ×100 magnification, 17 used
×100 and one lower magnification, three used ×100 and
two lower magnifications, and one used ×63 and one
lower magnification.
Experience With FISH
The total number of utilizations of probes with FISH
by the 26 laboratories during 1995 was 4,304 (median
108, range 0–728). Of these utilizations, 2,528 (58.7%)
employed LS probes, 1,440 (33.5%) employed alphoids,
and 336 (7.8%) employed CSPs. Of the utilizations,
2,420 (56.2%) involved peripheral blood, 1,189 (27.6%)
involved amniotic fluid, and 695 (16.1%) involved bone
marrow. During 1995, 21 participants used CSPs, 24
used alphoids, and 23 used LS probes (Table I). Only
one laboratory used no FISH for clinical evaluations in
1995. Of the remaining participants, 20 used FISH on
Multicenter Study With Interphase FISH
TABLE I. Experience by Laboratory*
Probe type
Utilizations
0
1–20
21–100
>100
Tissue type
CSP
Alph
LS
AF
BM
PB
5
18
3
—
2
10
10
4
3
4
8
11
6
15
4
1
7
14
3
2
3
4
13
6
*CSP, chromosome-specific paints; Alph, alpha satellites; LS, locus specific; AF, amniotic fluids; BM, bone marrows; PB, peripheral bloods.
amniotic fluid, 19 used FISH on bone marrow, and 23
used FISH on peripheral blood (Table I). Only 9 of 26
participants used the CEP-X/Y probes at least once in
1995.
Analytical Sensitivity
Metaphases and interphases from chromosomally
normal males and females were examined in Parts I, II,
and III of the project. Patterns of signals in cells from
each part of the project were pooled (Table II) in three
categories: XY, XX, and non-XY/non-XX (e.g., X or XXY
for males and X or XXX for females). The percentages
of XY cells (i.e., analytical sensitivities for male cells)
were very similar whether the data originated from one
male (specimens 3 and 9) or from 26 different males
(specimen 1). Similarly, the percentages of XX cells
(i.e., analytical sensitivities for female cells) were very
similar whether the 26 pooled samples originated from
one female (specimens 4 and 10) or from 26 different
females (specimen 2).
By using the CEP-X/Y cocktail of two probes, overall,
participants scored 0 of 3,900 metaphases and 5 of
13,000 interphases from normal males as XX, but they
scored 46 of 3,900 metaphases and 306 of 13,000 interphases as non-XY/non-XX (specimens 1, 3, and 9).
321
Overall, participants scored 4 of 3,900 metaphases and
6 of 13,000 interphases from normal females as XY, but
they scored 120 of 3,900 metaphases and 587 of 13,000
interphases as non-XY/non-XX (specimens 2, 4, and
10).
The percentage of XX cells ranged from 0 to 0.08 in
the pooled data from chromosomally normal males.
The corresponding range for XY cells in the pooled data
from chromosomally normal females was 0–0.31. The
percentage of non-XY/non-XX signals in metaphases or
interphases from male specimens did not exceed 2.4,
but all percentages of such signals in female specimens
equaled or exceeded 2.6 (Table II).
Artificial Chimeras
Metaphases and interphases from three artificial
chimeras were examined in Part II (specimens 5–7)
and in Part III (specimens 11, 8, and 12) of the project.
Data from each participant were also pooled in three
categories: XY, XX, and non-XY/non-XX (Table III).
Percentages of cells in each category differed among
the artificial chimeras for metaphases and for interphases. However, percentages of cells in each category
were very similar between metaphases and interphases and between repeat examinations of the same
artificial chimera.
Metaphases and interphases from the artificial chimeras were collected in samples of differing sizes in
Part III. Table IV shows participants that failed to detect XY cells at least twice in those samples. With the
frequency of the XY cells at about 44% or 19%, 26 of 26
participants detected at least two XY cells in 50 metaphases or interphases. However, with the frequency of
XY cells at about 8%, 4 of 26 participants did not find
two XY cells in 50 metaphases, and 3 of 26 did not find
two in 50 interphases.
TABLE II. Signal Patterns in Metaphases and Interphases From Normal Males and Females
Part/specimen no.
(experimental setup)
Metaphasesa
Interphasesb
Signal
Mean %
(Range)
Mean %
(Range)c
I/no. 1 (26 ‘‘local males’’)
XY
XX
Non-XY, non-XX
97.8
0.0
2.2
(86–100)
(0–0)
(0–14)
97.6
0.0
2.4
(85.0–100.0)
(0.0–0.5)
(0.0–15.0)
II/no. 3 (26 laboratories: 1 male)
XY
XX
Non-XY, non-XX
99.2
0.0
0.8
(90–100)
(0–0)
(0–10)
97.5
0.1
2.4
(88.0–100.0)
(0.0–1.0)
(0.0–7.5)
III/no. 9 (repeat of no. 3)
XY
XX
Non-XY, non-XX
99.5
0.0
0.5
(94–100)
(0–0)
(0–6)
97.9
0.0
2.1
(86.0–100.0)
(0.0–0.0)
(0.0–14.0)
I/no. 2 (26 ‘‘local females’’)
XY
XX
Non-XY, non-XX
0.0
96.2
3.8
(0–0)
(82–100)
(0–100)
0.0
96.2
3.8
(0.0–1.0)
(86.5–100.0)
(0.0–13.5)
II/no. 4 (26 laboratories: 1 female)
XY
XX
Non-XY, non-XX
0.0
97.4
2.6
(0–0)
(92–100)
(0–8)
0.0
95.4
4.6
(0.0–0.5)
(86.0–100.0)
(0.0–14.0)
III/no. 10 (repeat of no. 4)
XY
XX
Non-XY, non-XX
0.3
96.8
2.8
(0–8)
(86–100)
(0–8)
0.1
94.1
5.8
(0.0–2.0)
(69.0–100.0)
(0.0–31.0)
a
c
Fifty metaphases/specimen/participant.
Two hundred interphases/specimen/participant in Parts I and II; 100 in Part III.
Mean percentage was calculated from pooled observations and was rounded to the nearest 0.1%. Ranges include actual observations of individual
participants expressed as percentages.
b
c
322
Dewald et al.
TABLE III. Signal Patterns in Metaphases and Interphases From Three Artificial Chimeras
XY
Specimen no. (part;
experimental setup)
XX
Non-XY, non-XX
Mean %a
(Range)
Mean %a
(Range)
Mean %a
(Range)
Metaphases
No. 5 (part II)
No. 11 (part III; repeat of no. 5)
No. 6 (part II)
No. 8 (part III; repeat of no. 6)
No. 7 (part II)
No. 12 (part III; repeat of no. 7)
43.5
43.4
19.1
20.1
8.1
8.5
(30–56)
(26–54)
(12–24)
(12–32)
(2–18)
(0–16)
54.7
55.8
79.1
78.3
88.8
90.0
(44–68)
(44–74)
(66–88)
(66–88)
(62–96)
(82–98)
1.8
0.8
1.8
1.6
3.1
1.5
(0–10)
(0–6)
(0–8)
(0–10)
(0–20)
(0–6)
Interphasesc
No. 5 (part II)
No. 11 (part III; repeat of no. 5)
No. 6 (part II)
No. 8 (part III; repeat of no. 6)
No. 7 (part II)
No. 12 (part III; repeat of no. 7)
44.0
42.0
19.0
18.8
7.6
7.4
(35.5–52.5)
(32.0–60.0)
(8.5–26.0)
(6.0–32.0)
(3.5–13.0)
(2.0–17.0)
53.1
55.0
77.8
77.0
88.4
88.2
(47.0–63.0)
(40.0–68.0)
(71.5–86.0)
(63.0–94.0)
(82.5–94.0)
(69.0–100.0)
2.9
3.0
3.2
4.2
4.0
4.4
(0.0–6.5)
(0.0–12.0)
(0.0–9.5)
(0.0–11.0)
(0.0–9.5)
(0.0–11.0)
b
a
Mean percentage was calculated from pooled observations and was rounded to the nearest 0.1%. Ranges include actual observations of individual
participants expressed as percentages.
b
Fifty metaphases/specimen/participant.
c
Two hundred interphases/specimen/participant in part II; 100 in part III.
Normal Ranges
Normal ranges of the percentages of XY, XX, and
various non-XY/non-XX patterns of signals were constructed to assess the consistency of data reported from
different parts of this project. Nearly all samples of 50
metaphases from the 26 local males (specimen 1) or
from repeat samples from the target male (specimen 9)
fell within the normal (i.e., target) ranges developed
from the target male (specimen 3), as shown in Table V.
Nearly all repeat samples of 100 interphases from the
target male (specimen 9) and every sample of 200 interphases from the 26 local males (specimen 1) fell
within the normal ranges. Most samples of metaphases
or interphases from the 26 local females (specimen 2)
and from repeat samples of the target female (specimen
10) fell within the normal ranges developed from the
target female (specimen 4). Only 3 of 26 repeat samples
of 50 metaphases exceeded one or both target ranges
for XY and XX in the artificial chimeras (Table VI).
However, 8 or 9 of 26 repeat samples of 100 interphases
from the artificial chimeras exceeded one or both target
ranges.
Relating Analytical Sensitivity and Numbers
of Observations
A statistical approach based on the binomial distribution was used to relate analytical sensitivity to the
TABLE IV. Number of Laboratories (Out of 26) Failing to
Observe at Least Two XY Signals in Samples From
Artificial Chimeras
Observeda
Sample size (no. cells/specimen)
% XY
% XX
5
10
15
20
50
100
Metaphases
43.5
19.1
8.1
54.7
79.1
88.8
3
19
24
2
8
19
0
3
19
0
5
10
0
0
4
ND
ND
ND
Interphases
44.0
19.0
7.6
53.2
77.8
88.4
7
16
25
0
11
21
0
2
16
0
3
16
0
0
3
0
0
0
a
Data from part II. ND, not done.
numbers of observations required to assay for potential
mosaicisms or chimerisms. Examples of the approximate numbers of observations required to assure detection (with high probability) of possible second cell
lines (clones) of various frequencies are presented in
Table VII. For example, with 74 cells scored and an
analytical sensitivity of 95%, potential second cell lines
TABLE V. Percentage of Laboratories Outside the Target
Ranges for Metaphases and Interphases From Males
and Females
Laboratories outside
target range (%)
Signal
Males
Metaphases
XY
X
XX
XXY
XYY
Interphases
XY
X
XX
XXY
XYY
Females
Metaphases
XX
X
XXX
XY
XXY
Interphases
XX
X
XXX
XY
XXY
a
Target rangea
(part II)b
Part III
(repeat)c
Part I
(26 local subjects)b
96–100
0–2.0
0
0
0
0.0
3.8
0.0
3.8
0.0
11.5
11.5
0.0
7.7
0.0
92.5–100
0–4.0
0–1.0
0–2.0
0–1.0
0.0
7.7
0.0
0.0
3.8
0.0
0.0
0.0
0.0
0.0
92–100
0–4.0
0–4.0
0
0
3.8
7.7
0.0
3.8
0.0
15.4
11.5
3.8
0.0
3.8
89.5–99.5
0.5–8.0
0–3.0
0
0
11.5
7.7
11.5
7.7
3.8
7.7
0.0
3.8
3.8
7.7
Target ranges equal approximately 92% of the observed distributions.
Based on 50 metaphases and 200 interphases/specimen/participant. Part
II: males, specimen 3; females, specimen 4.
c
Based on 50 metaphases and 100 interphases/specimen/participant.
b
Multicenter Study With Interphase FISH
TABLE VI. Percentage of Laboratories Exceeding the Target
Ranges for XX and/or XY Signals in Metaphases and
Interphases From Artificial Chimeras
% XX
% XY
Repeats
outside one
or both rangesb
48–66
70–86
80–96
32–50
12–26
2–16
11.5
11.5
11.5
47.5–60.0
72.0–84.0
83.5–92.5
36.5–49.5
14.0–25.0
4.0–12.0
34.6
30.8
34.6
Target rangea
Sp
Metaphases
5
6
7
Interphases
5
6
7
a
Target ranges equal approximately 92% of the observed distributions in
part II.
b
Based on 50 metaphases and 100 interphases/specimen/participant.
encountered with a frequency of ù20% could be distinguished from the background inherent in FISH. The
values in Table VIII illustrate the probability of observing various numbers (and percentages) of potential second cell lines in samples of different sizes at different
analytical sensitivities. For example, with 50 cells
scored and an analytical sensitivity of 90%, the probability of observing ù10 cells from a second cell line is
0.025, the probability of observing ù11 cells is 0.009,
and the probability of observing ù9 cells is 0.058. Thus,
the chances of observing examples of any potential second cell line in an assay with N cells scored can be used
to distinguish a clone from the background inherent in
FISH. An arbitrary but typical probability cutoff is
0.05, but, often, the actual probabilities in a range up to
0.1 may be informative.
Workload Recording
Twenty-one participants provided complete timings
of the work required to follow the manufacturer’s protocol (Table IX). The mean time to process a slide, to
then score 50 metaphases, and to capture two documentary images was 27.1 minutes. The corresponding
mean time for 200 interphases was 28.6 minutes. To
make the required observations, only 1 of 21 participants processed a backup slide for specimen 6; none
TABLE VII. Approximate Numbers of Metaphases or
Interphases Required to Identify Mosaicisms or Chimerisms at
95% or 99% Power by Using Probes With Analytical
Sensitivities of 90%, 95%, or 99%
95% Powera
99% Powera
Analytical
sensitivity
Analytical
sensitivity
Frequency
(%) of second
cell lineb
90%
95%
99%
90%
95%
99%
ù50
ù40
ù30
ù20
ù10
ù5
11
19
39
133
ID
ID
7
12
20
46
291
ID
5
7
11
19
54
171
18
30
61
203
ID
ID
12
19
33
74
447
ID
8
12
18
34
92
282
a
Power in this context is the probability that the test rejects the null
hypothesis, HO, when, in fact, HO is false.
b
Frequency 4 1 − pa. See Materials and Methods, Calculations. ID, indistinguishable from background.
323
processed a backup for specimen 7. On average, participants took another 5.5 minutes to prepare solutions
for processing slides.
DISCUSSION
This project was designed to simulate actual studies
with FISH in clinical practice, including known specimens (Part I), unknown specimens (Parts II and III),
and abnormal specimens (artificial chimeras; Parts II
and III). The study focused on the sex chromosome constitution of metaphases and interphases, using differently colored probes for the pericentromeric segment of
the X and band q12 of the Y. In controlled trials, others
have demonstrated the reproducibility [Bryant et al.,
1995] and accuracy [Patil et al., 1995] of the probes. In
the present investigation, participants followed the
manufacturer’s protocols without appreciable local
modification. All participants supplied the specified observations for each segment of the three-part project,
and 21 of 26 participants provided complete measurements of the time expended on FISH. Consequently,
the results from each participating laboratory were
considered comparable samples for statistical analysis.
Part I of the project provided participants with an
opportunity to become familiar with the manufacturer’s instructions for using the probes and for scoring
signals [Schad and Dewald, 1995]. In Part I, participants were least experienced with the probes, and each
examined metaphases and interphases in two specimens from their local practices. Nevertheless, in neither Part I nor in subsequent parts of the project were
relationships apparent between results and participants’ equipment for visualizing signals from probes or
participants’ experiences with probes in 1995.
Analytical sensitivity is the proportion of all signal
patterns that matches the expectation for metaphases
and interphases, and, as such, it is a statistic that accounts for all sources of variation: biological, technological, and observational. In this study, analytical sensitivities were consistently high whether metaphases
or interphases from a male or a female were examined
twice by each of the 26 participants or whether each
participant examined a different male and female
(Table II). In the examinations of cells from chromosomally normal males and females, opposite sex patterns of signals occurred in ø0.3% of 1,300 metaphases
and in ø0.1% of 2,600 interphases, but the percentages
of non-XY/non-XX signals were 0.5–2.4% for males and
2.6–5.8% for females. These data suggest that the CEPX/Y probe cocktail could be expected to detect lower
frequencies of chimerism than sex chromosome mosaicism. Nevertheless, when they were used according to
the manufacturer’s instructions, the probes for X and Y
proved to be robust and reliable, even when participants had various experiences with probes (Table I)
and unique combinations of equipment for viewing signals from FISH.
Three artificial chimeras served as unknown specimens for this project. The relative proportions of XY
and XX signals were strikingly different among the
three artificial chimeras for metaphases and for interphases (Table III). Two separate sets of examinations
324
Dewald et al.
TABLE VIII. Probability of Observing Specified Numbers of a Second Cell Line in Tests With Different Analytical Sensitivities
AS 4 90%a
Cells
scored
50
100
200
300
400
500
1,000
a
Abnormal
cells
P
9
10
11
12
13
15
16
17
18
19
20
21
27
28
29
30
31
32
33
34
35
39
40
41
42
43
44
45
46
47
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
0.058
0.025
0.009
0.003
0.001
0.073
0.040
0.021
0.010
0.005
0.002
0.001
0.067
0.043
0.027
0.016
0.010
0.005
0.003
0.002
0.001
0.055
0.038
0.025
0.017
0.011
0.007
0.004
0.002
0.001
0.060
0.044
0.031
0.022
0.015
0.010
0.007
0.004
0.003
0.002
0.001
0.062
0.046
0.034
0.025
0.018
0.013
0.009
0.006
0.004
0.003
0.002
0.001
0.053
0.043
0.035
0.028
0.022
0.017
0.013
0.010
0.008
0.006
0.005
0.003
0.003
0.002
0.001
AS 4 95%
AS 4 96%
Abnormal
cells
P
5
6
7
8
9
9
10
11
12
13
0.104
0.038
0.012
0.003
0.001
0.063
0.028
0.011
0.004
0.001
15
16
17
18
19
20
21
AS 4 97%
AS 4 98%
P
Abnormal
cells
Abnormal
cells
P
5
6
7
8
0.049
0.014
0.004
0.001
5
6
7
8
0.063
0.017
0.004
0.001
7
8
9
10
11
12
0.106
0.048
0.019
0.007
0.002
0.001
6
7
8
9
10
0.078
0.044
0.024
0.012
0.006
0.003
0.001
13
14
15
16
17
18
0.060
0.031
0.015
0.007
0.003
0.001
22
23
24
25
26
27
28
0.049
0.029
0.017
0.009
0.005
0.003
0.001
18
19
20
21
22
23
24
27
28
29
30
31
32
33
34
35
0.073
0.048
0.031
0.019
0.011
0.007
0.004
0.002
0.001
33
34
35
36
37
38
39
40
41
42
62
63
64
65
66
67
68
69
70
71
72
73
AS, analytical sensitivity; P, probability.
Abnormal
cells
AS 4 99%
P
Abnormal
cells
P
3
4
5
0.078
0.018
0.003
2
3
4
0.089
0.014
0.002
0.081
0.031
0.011
0.003
0.001
5
6
7
8
0.051
0.015
0.004
0.001
3
4
5
6
0.079
0.018
0.003
0.001
10
11
12
13
14
15
0.081
0.040
0.018
0.008
0.003
0.001
8
9
10
11
12
0.049
0.020
0.007
0.003
0.001
5
6
7
8
0.052
0.016
0.004
0.001
0.059
0.034
0.019
0.010
0.005
0.002
0.001
14
15
16
17
18
19
20
0.071
0.039
0.020
0.010
0.005
0.002
0.001
10
11
12
13
14
15
0.082
0.041
0.019
0.008
0.003
0.001
6
7
8
9
10
0.083
0.033
0.011
0.004
0.001
23
24
25
26
27
28
29
30
0.054
0.034
0.020
0.011
0.006
0.003
0.002
0.001
18
19
20
21
22
23
24
0.060
0.035
0.020
0.010
0.005
0.003
0.001
13
14
15
16
17
18
0.062
0.033
0.016
0.008
0.003
0.001
8
9
10
11
12
0.050
0.021
0.008
0.003
0.001
0.066
0.045
0.030
0.020
0.012
0.008
0.005
0.003
0.002
0.001
28
29
30
31
32
33
34
35
0.049
0.031
0.020
0.012
0.007
0.004
0.002
0.001
22
23
24
25
26
27
28
0.050
0.031
0.018
0.010
0.005
0.003
0.001
15
16
17
18
19
20
21
0.081
0.047
0.026
0.013
0.007
0.003
0.001
9
10
11
12
13
14
0.067
0.031
0.013
0.005
0.002
0.001
0.051
0.038
0.028
0.021
0.015
0.011
0.007
0.005
0.003
0.002
0.002
0.001
51
52
53
54
55
56
57
58
59
60
61
0.049
0.036
0.026
0.018
0.012
0.008
0.006
0.004
0.002
0.002
0.001
39
40
41
42
43
44
45
46
47
48
0.062
0.044
0.030
0.020
0.014
0.009
0.006
0.003
0.002
0.001
28
29
30
31
32
33
34
35
0.051
0.033
0.021
0.013
0.008
0.004
0.002
0.001
15
16
17
18
19
20
21
0.082
0.048
0.026
0.014
0.007
0.003
0.001
Multicenter Study With Interphase FISH
TABLE IX. Workload for CEP-X/Y Probe (Min:Sec/Specimen):
Twenty-One Laboratories
Task
Mean
S.D.
Max
Min
Denature specimen DNA
Denature probe
In situ hybridization
Posthybridization washes
Analyze interphase
Analyze metaphase
Total interphase
Total metaphase
Solution preparation
4:49
3:06
1:14
2:04
17:21
15:50
28:34
27:03
5:30
2:49
1:41
0:33
0:49
9:20
7:46
9:28
8:30
2:32
10:11
6:58
2:33
3:48
50:01
45:40
59:03
50:29
10:56
1:25
0:33
2:25
0:50
5:34
6:41
15:50
12:27
2:02
produced essentially identical proportions of XY and
XX signals for each of the three different artificial chimeras. Signal patterns from the repeat study of each
artificial chimera were collected in the sample sizes
shown in Table IV. No samples of ø50 observations
from all three of the artificial chimeras permitted a
high level of confidence in detecting at least two XY
signals, the minimum that might be needed to document a second cell line, or ‘‘clone,’’ in an otherwise homogeneous array of XX signals [ISCN, 1995].
All participants scored signals in 50 metaphases and
in 200 interphases from the same chromosomally normal male (specimen 3) and female (specimen 4). All
except the high and the low frequencies of XY, XX and
the various non-XY/non-XX signals in the 26 samples
were used to construct normal (i.e., target) ranges.
These ranges were used for comparisons with the 26
different chromosomally normal males (specimen 1)
and females (specimen 2) studied by individual participants (Table V). These ranges were also used for comparisons with smaller, repeat samples, which were examined blindly, of the same two specimens from which
the normal ranges had been developed. It is important
to note that data from specimens 1 and 2 were collected
first and that the 52 different specimens offered an
opportunity for variation that was not present in other
parts of the project. Nevertheless, nearly all frequencies of the various signals from metaphases or interphases of both comparison groups fell within their respective ranges. Consequently, we believe that these
normal ranges establish reasonable cutoffs for metaphases and interphases from chromosomally normal
males and females examined with CEP-X/Y. Because
they were established by 26 different laboratories,
these normal ranges represent more realistic estimates
of the performance of the probes in clinical practice
than might be generated by one or by a small number
of specifically trained and highly experienced testing
sites.
Each segment of the project involved metaphases, for
which clinical guidelines have been developed, and involved interphases, for which guidelines have not been
developed [American College of Medical Genetics,
1996]. Agreement between data from metaphases and
interphases was very high, although only 25% or 50%
as many metaphases as interphases were examined
per sample (Tables II and III). Nevertheless, metaphases from chromosomally normal males and females
and from the artificial chimeras consistently exhibited
lower percentages of non-XY/non-XX signals than in-
325
terphases. Furthermore, with increasing proportions of
XX signals in the artificial chimeras, the proportion of
non-XY/non-XX signals increased more in interphases
than in metaphases. These data suggest that, to
achieve equivalent levels of confidence, assessments of
interphases with FISH would require more observations than assessments of metaphases.
Presently, no agreement exists about the numbers of
observations needed for clinical studies with interphase FISH [American College of Medical Genetics,
1996]. In this project, samples of 50 interphases allowed about 88% of participants to detect at least two
interphases with XY signals in an artificial chimera in
which the frequency of nuclei with XY was about 8%
(Table IV). Only 65–70% of samples of 100 interphases
fell within the normal range for the relative proportions of XY and XX signals for each of three artificial
chimeras (Table VI). However, all samples of 200 interphases from 26 different males and 92–100% of
those from 26 different females fell within normal
ranges established from one male and one female specimen studied by all participants (Table V). These data
suggest that to approach the confidence levels typically
associated with standard cytogenetics for detecting cell
lines occurring at frequencies of ù10% and to be confident of distinguishing cell lines from backgrounds,
including anomalous signals, would require samples of
ù200 interphases.
Frequencies of non-XY signals in metaphases and
interphases from normal males were low (Table II).
The same was the case for non-XX signals in metaphases and interphases from normal females. Such
non-XY/non-XX signals comprise an inherent background of anomalous signals (e.g., the normal or target
ranges; Table V) from which any potential second cell
line (clone) would have to be distinguished. All such
non-XY/non-XX signals, as illustrated in Table II, occur
in metaphases and interphases from normal individuals with a frequency approximating one minus the analytical sensitivity. Logically, then, the closer the analytical sensitivity of any probe approaches unity, the
more a relatively few anomalous signals in some number of observations might indicate the presence of a
second cell line, a true mosaicism or chimerism, in an
unknown specimen. Conversely, the further the analytical sensitivity departs from unity, the less likely it
is that a relatively few anomalous signals might indicate true mosaicism or chimerism.
Consequently, a statistical approach based on the binomial distribution was used to relate analytical sensitivity of a probe to the numbers of observations required to assay for mosaicisms or chimerisms. The values in Table VII illustrate for 95% and 99% power the
striking effect analytical sensitivity has on the number
of observations necessary to assure detection of second
cell lines: lower analytical sensitivities require much
larger numbers of observations. The values in Table
VIII illustrate the profound effect of analytical sensitivity on the chance of observing the specific numbers
of abnormal cells required to establish statistically significant levels of second cell lines: lower analytical sensitivities require relatively much larger numbers of abnormal observations. These effects demonstrate the
326
Dewald et al.
crucial importance of maintaining a consistent, locally
determined analytical sensitivity for each probe used
in clinical evaluations to assure the capacity to identify
significant mosaicisms or chimerisms accurately and to
manage workload.
The effort associated with FISH was estimated by
timing the work required for mixing two solutions, processing the prepared slides, and observing (and documenting) 50 metaphases and 200 interphases. Notable
variation was evident in the complete timings of workload (Table IX). For metaphases and interphases, the
maximum and minimum total times differed about
fourfold. Some of the variation may be ascribed to inexperience with timing work and to inaccurate timing
of many short steps. Nevertheless, we believe that the
bulk of the variation can be attributed to differing efficiencies between laboratories. With average times for
the various components of processing and analyzing
slides, it should be possible for laboratories to identify
misplaced effort, to compute an average cost of the labor associated with FISH, and to estimate the staff
needed for studies with probes.
Spectrum CEP X/Y dual color probe kit for enumeration of XX and XY
cells in opposite-sex bone marrow transplant patients: A multicenter
clinical validation study. Am J Hum Genet 57(A109):607.
Dewald GW, Schad CR, Christensen ER, Tiede AL, Zinsmeister AR, Spurbeck JL, Thibodeau S, Jalal JM (1993a): The application of fluorescent
in situ hybridization to detect Mbcr/abl fusion in variant Ph chromosomes in CML and ALL. Cancer Genet Cytogenet 71:7–14.
Dewald GW, Schad CR, Christensen ER, Law ME, Zinsmeister AR, Stalboerger PG, Jalal SM, Ash RC, Jenkins RB (1993b): Fluorescent in situ
hybridization with X and Y chromosome probes for cytogenetic study on
bone marrow cells after opposite sex transplantation. Bone Marrow
Transplant 12:149–154.
Dewald GW, Stallard R, Bader PI, Chen K, Zenger-Hain J, Harris CJ,
Higgins R, Hirsch Hsu WT, Johnson E, Kubic V, Kurczynski TW,
Malone JM, McCorquodale DJ, Meilinger K, Meisner LF, Moore JW,
Schwartz S, Siembieda S, Storto PD, Vance G, Van Tuninen P, Wiktor
A, Yung JF (1996): Toward quality assurance for metaphase FISH: A
multicenter experience. Am J Med Genet 65:1–8.
Dewald GW, Brothman AR, Butler MG, Cooley LD, Patil SR, Saikevych IA,
Schneider NR (1997): Pilot studies for proficiency testing using fluorescence in situ hybridization with chromosome-specific DNA probes.
Arch Pathol Lab Med 121:359–367.
ISCN (1995): ‘‘An International System for Human Cytogenetic Nomenclature.’’ Mitelman F (ed): Basel: S. Karger.
Jenkins RB, Le Beau MM, Kraker WS, Borell TJ, Stalboerger PG, Davis
EM, Penland L (1992): Fluorescence in situ hybridization: A sensitive
method for trisomy 8 detection in bone marrow specimens. Blood 79:
3307–3315.
ACKNOWLEDGMENTS
Koska R (1997): CEP 8, CEP 12 and CEP X/Y DNA probe kits first FDA
cleared products utilizing FISH technology. Vysions 2:6–7.
This work was supported in part by project MCJ551004-04 from the Maternal and Child Health Bureau
(Social Security Act, Title V; Health Resources and Services Administration, Department of Health and Human Services) and by Vysis, Inc. (Downers Grove, IL),
which donated the DNA probes.
Lu PY, Hammitt DG, Zinsmeister AR, Dewald GW (1994): Dual color fluorescence in situ hybridization to investigate aneuploidy in sperm from
33 normal males and a man with a t(2;4;8)(q23;q27;p21). Fertil Steril
62:394–399.
REFERENCES
Schad CR, Dewald GW (1995): Building a new clinical test for fluorescence
in situ hybridization. Appl Cytogenet 21:1–4.
American College of Medical Genetics (1996): ‘‘Standards and Guidelines:
Clinical Genetic Laboratories, Supplement 1.’’ Bethesda, MD: The
American College of Medical Genetics Laboratory Practice Committee.
Spurbeck JL, Zinsmeister AR, Meyer KJ, Jalal SM (1996): The dynamics of
chromosome spreading. Am J Med Genet 61:387–393.
Bryant E, Patil SR, Theil K, Seelig S, Hsu P (1995): Reproducibility of
Patil SR, Bryant E, Theil K, Seelig S, Hsu P (1995): Accuracy of spectrum
CEP X/Y dual color probe kit for detection of XX and XY cells in opposite-sex bone marrow transplant patients: A multicenter clinical validation study. Am J Hum Genet 57(A74):397.
Trask B (1991): Fluorescence in situ hybridization: Applications in cytogenetics and gene mapping. Technical Focus 7:149–154.
Документ
Категория
Без категории
Просмотров
4
Размер файла
61 Кб
Теги
using, investigation, fluorescence, multicenter, hybridization, interphase, probes, chromosome, situ
1/--страниц
Пожаловаться на содержимое документа