Clinical Chemistry 46: 1106-1113, 2000;
(Clinical Chemistry. 2000;46:1106-1113.)
© 2000 American Association for Clinical Chemistry, Inc.
Description of a Computer Program to Assess Cancer Antigen 15.3, Carcinoembryonic Antigen, and Tissue Polypeptide Antigen Information during Monitoring of Metastatic Breast Cancer
György Sölétormos1,2,3,a and
Vibeke Schiøler1
1
Department of Clinical Biochemistry, Herlev Hospital, University of Copenhagen, DK-2700 Copenhagen, Denmark and
2
Oncology, Herlev Hospital, University of Copenhagen, DK-2700 Copenhagen, Denmark.
3
Department of Clinical Biochemistry, Hillerød Hospital, DK 3400 Hillerød, Denmark.
a Address correspondence to this author at: Department of Clinical Biochemistry, Hillerød Hospital, Helsevej 2, DK 3400 Hillerød, Denmark. Fax 45-4824-0067; e-mail geso{at}fa.dk
 |
Abstract
|
|---|
It is time-consuming to process and compare the clinical and marker
information registered during monitoring of breast cancer patients. To
facilitate the assessment, we developed a computer program for
interpreting consecutive measurements. The intraindividual biological
variation, the analytical precision profile, the cutoff limit, and the
detection limit for each marker are entered and stored in the program.
The assessment procedure for marker signals considers the analytical
and biological variation of the applied markers. The software package
contains a database that can store the interpretation of the
measurements as evaluation codes together with patient demographics,
information about treatment type, dates for treatment periods, control
periods, and evaluation codes for clinical activity of disease. The
consecutive concentrations for a patient are imported temporarily into
the program from outside sources and presented graphically. Marker
concentrations to be compared are selected with the computer mouse and
the significance of the difference is calculated by the program. The
program has an option for calculating the lead time of marker signals
vs clinical information. The program facilitates the monitoring of
individual breast cancer patients with tumor marker measurements. It
may also be implemented in trials investigating the utility of
potential new markers in breast cancer as well as in other
malignancies.
|
Introduction
|
|---|
Many compounds have been studied in the search for tumor markers
that rapidly and reliably reflect treatment response of the individual
breast cancer patient or give an early prediction of recurrence or new
metastases (1)(2)(3). Longitudinal monitoring studies
necessitate a standardized set of criteria for marker assessment
(4)(5). However, calculation of the significance
of a change in concentrations is very time-consuming if performed
manually (6)(7). To facilitate this process, we
have developed a computer program based on our model systems for tumor
marker monitoring where the significance of a change is related to the
normal inherent intraindividual biological variation
(CVi) and the total analytical variation
(CVa) of the applied markers and assay methods,
respectively (5)(6)(7). The program performs all calculations
needed to interpret consecutive measurements and matches and compares
marker and clinical information.
|
Materials, Methods, and Results
|
|---|
Patients
We previously have compared the ability of cancer antigen 15.3 (CA
15.3),1
carcinoembryonic antigen (CEA), tissue polypeptide antigen
(TPA), and routine clinical procedures to signal changed disease
activity among 204 metastatic breast cancer patients monitored during
first-line chemotherapy and subsequent follow-up (7). The
majority of the calculations were performed without computer
assistance. Here we provide data from 59 of the patients randomized to
epirubicin treatment. All data were calculated, processed, and
presented by use of the generated computer program.
Tumor markers
The CA 15.3, CEA, and TPA concentrations were determined with
commercial radioimmunoassays: CA 15.3 (International CIS), CEA
(Kabi Pharmacia), and TPA (Byk-Sangtec). Blood specimens for marker
analysis were sampled every 3–4 weeks according to scheduled time
points. Each specimen was analyzed for CA 15.3, CEA, and TPA. The
specimens were analyzed consecutively, and each specimen from an
individual patient was analyzed in a separate assay run.
Overview of the computer program
The computer program was written in Microsoft Visual Basic 3.0
under Windows 95/Windows 98. The organization, options, functions, and
workflow of the program are summarized in Table 1
.
Support section
Test catalog.
The unit of measurement, the measuring range,
the detection limit, and the cutoff limit provided in the package
insert are entered into the "test catalog" together with the values
for within-subject biological variation (not including analytical
variation): 6.2% for CA 15.3, 9.3% for CEA, and 28.3% for TPA
(5). The analytical imprecision represented as a precision
profile is determined by each department. The measuring range is
divided into 20 intervals with the finest divisions where the precision
profile is steepest. For each interval, the values for concentration
and the corresponding analytical imprecision are entered into the
program in a tabular format. A graphical presentation of the precision
profile data with the detection and cutoff limit is optional. The
imprecision corresponding to a selected concentration interval is shown
by pointing at the precision profile with the mouse. The test
catalog may hold information for up to 100 different marker
assays and may be updated as new markers are implemented or measuring
methods are improved or changed.
Evaluation codes catalog.
The codes for clinical as well as
marker assessment have to be entered into the "evaluation codes
catalog" section. The clinical assessment of disease status according
to WHO (8) comprises the following codes:
(a) disappearance of all known disease [complete response
(CR)]; (b) a decrease of 
50% in total tumor size
[partial response (PR)]; (c) an increase 25% in any
lesion [progressive disease (PD)]; (d) stable disease,
defined as neither response nor progression of disease [no change
(code NC)]. Patients dying within 4 weeks after initiation of therapy
were registered as having early death (code ED). Patients dying without
clinical signs of PD were registered as PD at death (code PDM).
There is no consensus regarding interpretation of consecutive marker
concentrations. Several criteria have been proposed
(6)(7)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34). The data suggest that
the criteria generated by Sölétormos and co-workers
(6)(7)(34), which considered the
cutoff value, the duration of the change, and adjusted the significance
of a difference to the background variation of the investigated marker,
performed better than criteria where these parameters were not
integrated into the assessment procedure (9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34). The high
reliability of the criteria proposed by Sölétormos and
co-workers (6)(7) has been substantiated in
computer-simulated models in which the different criteria from several
authors were compared (34). The criteria elaborated by
Sölétormos and co-workers
(6)(7)(34) as well as the associated
marker evaluation codes have been described in detail previously and
summarized below: (a) significant decrement from above to
below cutoff (complete marker response; CR); (b) significant
decrement from above cutoff to a lower value above cutoff (PR);
(c) significant increment from below to above cutoff or from
above cutoff to a higher value (PD); (d) marker
concentrations fluctuated and fulfilled neither response nor
progression criteria [no change above cutoff (code NCH) or no change
below cutoff (code NCL)]. Additionally, it may be necessary to have
codes for steady-state conditions if the tumor marker information
changes without a change in tumor status or vice versa [unchanged
status of complete response (code CRK) or unchanged status of partial
response (code PRK)].
Patient database section
Notebook.
The "notebook" contains
patient name, patient identification number (Pat.ID), patient
registration number in a trial (Regist.no.), and the location of the
patient. Table 2
shows a printout of a patient record. For each patient, the
notebook can hold information of four treatment and control periods.
The information to be entered is as follows: starting date of
treatment, end of treatment (treatment I to treatment IV), end of
control period (control period I to control period IV), and a remark
(I–IV) of the type of treatment. Each of these eight periods has a
table for storing four evaluation codes and the corresponding dates of
change for each of the markers as well as for the clinical status of
disease. The tables are accessed by clicking the treatment number
(I–IV) listed on the first screen of the notebook. It is possible to
select a printout of one or all eight evaluation tables. The program
also provides an option for combining information from more markers in
one evaluation code, which can be entered manually into the evaluation
tables.
Main section
Statistical methods.
A major function of the program is to
calculate the magnitude, the duration, and the significance of a change
in concentration. A change in two concentrations is significant if the
difference, expressed as a percentage of their mean value, exceeds
·
Z
(CVa2 +
CVi2) (35). The is a constant (for two
measurements). The Z-statistic depends on the probability
selected for significance and on whether the expected change is uni- or
bidirectional. At P = 0.05, Z = 1.65 if
the expected change is unidirectional (only one option, either an
increment or a decrement) and Z = 1.96 if
bi-directional (two options because it is unknown whether the
concentration will rise or fall). The CVa is the
analytical imprecision of the prechange concentration, and the
CVi is a population-based intraindividual
biological variation.
Graphical evaluation.
The tumor marker concentrations and the
corresponding dates of measurement are imported patient by patient into
the program from outside sources. The data are stored in a temporary
file, which is cleared when values for the next patient are imported or
when the program is closed. Up to four different markers for each
patient may be selected simultaneously for the evaluation procedure.
The sequential concentrations of the four markers and their respective
cutoff limits are provided on the same screen to facilitate an overview
(Fig. 1
; optional printout). Each graph can cover a time span of up to
3000 days, so it is possible to display patient data for more than 8
years.
View larger version (23K):
[in this window]
[in a new window]
|
Figure 1. Printout of a graphical overview of the different markers
and their concentrations determined in an individual patient.
MA8702 denotes the trial code. Registr.no.
39 denotes the patient registration number in the trial.
MUUMARBI denotes the name code. The lower part of the
screen provides information on how to proceed to the next step in the
graphical evaluation procedure (Fig. 2
). Activation of the
PrintScreen option provides a print of the marker
kinetics as presented above. Activation of the Exit
option returns the program to the main menu (Table 1
).
|
|
By pointing and clicking with the mouse, the operator can enlarge a
selected graph (Fig. 2
; optional printout). If the patient’s demographic data have
been registered in the notebook, the program connects these data with
the marker data, and markings on the graph indicating treatment
starting point, endpoint, and end of control periods are obtained. The
enlarged graph is presented with 500 days at a time and uses a scroll
button to scroll through the entire course of measurements. The
evaluation of data is performed on the enlarged graph. The first
measurement of the pair to be compared is selected by the mouse. The
point is color-marked; in addition, the concentration and the date of
measurement are shown on the screen. The second measurement is then
selected by the mouse, with the data also being shown on the screen
together with the number of days between the two measurements. After
the operator selects one-sided or two-sided testing, the program
performs all further steps necessary for the significance test.
Messages indicating "difference is significant" or
"difference not significant" are shown on the screen. The tested
significance levels are P <0.05, P <0.01, and
P <0.001.
View larger version (43K):
[in this window]
[in a new window]
|
Figure 2. Printout illustrating the graphical evaluation option of
CA 15.3 in an individual patient.
 , concentrations obtained during the monitoring period. •,
concentrations considered for the evaluation procedure. Tr I
end denotes the end of first-line treatment for metastatic
breast cancer (epirubicin) (7). Start date
denotes the sampling date of the concentration selected as the baseline
against which other concentrations are tested (ddmmyyyy). Eval.
date denotes the sampling date of the concentration tested for
significance compared with the baseline concentration. The significance
calculation was based on a two-side test because, theoretically, the
concentrations following the selected baseline concentrations could
either continue to decrease or increase. CV(a) denotes
the analytical imprecision of the selected baseline concentration;
CV(i) denotes the average intraindividual biological
background variation determined in a healthy female population
(5). PD denotes CA153 progression February
28, 1989. MUUMARBI denotes the name code for the
presented patient. Pat.ID denotes the patient
identification based on the date of birth, November 11, 1946.
Registr.no. denotes the registration number in the MA
8702 trial. R117 denotes the Department of Oncology
R117, Herlev Hospital, University of Copenhagen, Denmark. < and
> denote options for scrolling through those parts of the marker
concentrations that are outside the screen. Activation of the
PrintGraph option provides a printout of all
concentrations, including those outside the screen.
|
|
It is possible to select the evaluation code that describes the tumor
marker interpretation from the list of codes presented as a drop-down
list and to store the information (evaluation code and date of change
in concentration) if the patient’s demographic data have been entered
beforehand in the notebook.
Nongraphical evaluation.
An option for testing the difference
between two concentrations without importing all measurements for a
patient is also available. The marker considered is selected from a
drop-down list of available tumor markers (the "test list"), the
first and second measurements are entered, one-sided or two-sided
testing is selected, the "evaluate" option is clicked, and the
result of the significance test is shown on the screen. If the
patient’s demographics are in the notebook, it is possible to store
both the evaluation codes and the corresponding dates of change. A
printout is optional. A manual evaluation can also be performed for
patients not entered in the notebook, but then it is not possible to
store either the evaluation codes or the dates of change in marker
concentrations.
Match evaluation codes, list patients, and calculate lead
time.
The program can be used to match the tumor marker evaluation
codes with the clinical evaluation codes specified for each of the four
evaluation numbers as well as to summarize the results of the matching
process for all four evaluation numbers within a selected treatment or
control period (Table 3
). The program also has an option for finding patients with a
specified combination of marker and clinical evaluation codes and to
calculate the marker lead time for concordant events (CR-CR, PR-PR,
PD-PD; positive lead time, marker information precedes clinical
information; negative lead time, clinical information precedes marker
information; no lead time, the clinical and marker information is
obtained simultaneously). After the treatment or control period
and the evaluation number are selected, the marker evaluation code and
clinical evaluation code to search for are selected by clicking the
code from the drop-down code list. The program then finds the patients
with the specified combination of codes, identifies the corresponding
dates of change, calculates the lead times, and places the names, the
patient identification number, and the patient registration number on a
scroll list on the screen. A printout is optional (Table 4
).
View this table:
[in this window]
[in a new window]
|
Table 4. Printout of the lead timesa obtained for patients with CA 15.3 as well as a clinical evaluation code of PD.
|
|
|
Discussion
|
|---|
Serological tumor markers have been the subject of several
investigations suggesting a potential role in monitoring breast cancer.
Most commercially available assays are recommended for monitoring
purposes by the respective manufacturers. None, however, provides
guidelines for interpreting sequential concentrations measured during
therapy and follow-up.
Here we present a computer program that facilitates a standardized
interpretation of sequential concentrations. The program is based on
our previous studies of stochastic variation of tumor markers and on
the use of normal biological variation and analytical imprecision to
test the significance of a change in concentrations (5)(6)(7).
We have chosen to use a population-based mean intraindividual
biological variation in combination with the analytical variation in
the significance calculations because it seldom is possible to obtain
information about the intraindividual variation of a marker in the
single patient before disease, but any fraction (95th or 90th
percentile) instead of the mean may be entered as
CVi in the test catalog of the program if
preferred. The selected level should, however, depend on whether a high
sensitivity or a high specificity is considered important for the
monitoring situation in question.
Previous studies from our group have documented that the simple use of
cutoff limits for detecting unusual marker results for a particular
individual has little value because significant changes in
concentrations may occur within the reference range without any
tumor being present and because changes from within to outside the
reference range or vice versa are not necessarily significant
(5). However, if the cutoff limit is integrated together
with the magnitude and duration of a change into criteria for
assessment as recommended by Sölétormos and co-workers
(6)(7)(34), the marker information
becomes more reliable than information obtained using criteria
recommended by other authors (9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33).
Clinical trials on the utility of serological tumor markers may include
an enormous amount of data with hundreds of patients and measurements
of a large number of consecutive serum samples for each patient and
marker. Standardized evaluation codes and standardized procedures for
storing the codes are a prerequisite for computerized matching and
comparison of marker and clinical data. We have designed the tables for
storing and matching the evaluation codes with options for four
treatment and control periods and four evaluations during each of these
eight periods because both the activity of malignant disease and marker
concentrations often show dynamic changes over time (Table 2
).
In conclusion, the presented computer program may be implemented to
assess the performance of different tumor markers and to select the
most reliable marker in terms of responsive and progressive breast
cancer. When data that characterize the analytical and biological
variability are available, the program may also be used to monitor
other malignancies.
|
Acknowledgments
|
|---|
The computer program can be obtained by contacting the authors.
|
Footnotes
|
|---|
1 Nonstandard abbreviations: CA 15.3, cancer antigen 15.3; CEA, carcinoembryonic antigen; TPA, tissue polypeptide antigen; CR, complete remission; PR, partial remission; and PD, progressive disease. 
|
References
|
|---|
-
Chan DW, Sell S. Tumor markers. Burtis CA Ashwood ER eds. Tietz textbook of clinical chemistry, 3rd ed 1999:722-749 WB Saunders Philadelphia. .
-
Aziz KJ. Tumor markers: reclassification and new approaches to evaluation. Adv Clin Chem 1998;33:169-199.
[ISI][Medline]
[Order article via Infotrieve]
-
Duffy MJ. CA 15-3 and related mucins as circulating markers in breast cancer. Ann Clin Biochem 1999;36:579-586.
-
Tate H. Assessing tumour markers. Br J Cancer 1981;44:643-651.
[ISI][Medline]
[Order article via Infotrieve]
-
Sölétormos G, Schiøler V, Nielsen D, Skovsgaard T, Dombernowsky P. Interpretation of results for tumor markers on the basis of analytical imprecision and biological variation. Clin Chem 1993;39:2077-2083.
[Abstract]
-
Sölétormos G, Nielsen D, Schiøler V, Skovsgaard T, Winkel P, Mouridsen HT, Dombernowsky P. A novel method for monitoring high-risk breast cancer with tumor markers: CA 15.3 compared to CEA and TPA. Ann Oncol 1993;4:861-869.
[Abstract/Free Full Text]
-
Sölétormos G, Nielsen D, Schiøler V, Skovsgaard T, Dombernowsky P. Tumor markers cancer antigen 15.3, carcinoembryonic antigen, and tissue polypeptide antigen for monitoring metastatic breast cancer during first-line chemotherapy and follow-up. Clin Chem 1996;42:564-575.
[Abstract/Free Full Text]
-
. World Health Organization. WHO handbook for reporting results of cancer treatment 1979:22-27 WHO Geneva. .
-
Barak M, Steiner M, Finkel B, Abrahamson J, Antal S, Gruener N. CA-15.3, TPA and MCA as markers for breast cancer. Eur J Cancer 1990;26:577-580.
-
Bombardieri E, Pizzichetta M, Veronesi P, Seregni E,
Bogni A, Maffioli L. CA 15.3 determination in patients with breast
cancer: clinical utility for the detection of distant metastases. Eur J
Cancer 1993;29A:144–6..
-
Colomer R, Ruibal A, Genolla J, Rubio D, Del Campo JM, Bodi R, Salvador L. Circulating CA 15-3 levels in the postsurgical follow-up of breast cancer patients and in non malignant diseases. Br Cancer Res Treat 1989;13:123-133.
[ISI][Medline]
[Order article via Infotrieve]
-
Kallioniemi OP, Oksa H, Aaran RK, Hietanen T, Lehtinen M, Koivula T. Serum CA 15-3 assay in the diagnosis and follow-up of breast cancer. Br J Cancer 1988;58:213-215.
[ISI][Medline]
[Order article via Infotrieve]
-
Locker GJ, Mader RM, Braun J, Sieder AE, Marosi C, Rainer H, et al. New mucin-like cancer associated antigens (CAM 26, CAM 29 and CA 549) and a new proliferation marker (TPS) in patients with primary or advanced breast cancer. Oncology 1995;52:140-144.
[ISI][Medline]
[Order article via Infotrieve]
-
Markopoulos CJ, Gogas HJ, Alevizou-Terzaki BP, Gogas JG. CA 15-3 in the prediction of recurrence of breast cancer. Breast Dis 1994;7:1-5.
-
Molina R, Zanón G, Filella X, Moreno F, Jo J, Daniels M, et al. Use of serial carcinoembryonic antigen and CA 15.3 assays in detecting relapses in breast cancer patients. Breast Cancer Res Treat 1995;36:41-48.
[ISI][Medline]
[Order article via Infotrieve]
-
Nicolini A, Colombini C, Luciani L, Carpi A, Giuliani L. Evaluation of serum CA 15-3 determination with CEA and TPA in the post-operative follow-up of breast cancer patients. Br J Cancer 1991;64:154-158.
[ISI][Medline]
[Order article via Infotrieve]
-
Pectasides D, Pavlidis N, Gogou L, Antoniou F, Nicolaides C, Tsikalakis D, Fountzilas G. Clinical value of CA 15-3, mucin-like carcinoma-associated antigen, tumor polypeptide antigen, and carcinoembryonic antigen in monitoring early breast cancer patients. Am J Clin Oncol 1996;19:459-464.
[ISI][Medline]
[Order article via Infotrieve]
-
Ahlemann LM, Staab HJ, Anderer FA. Serial CEA determinations as an aid in postoperative therapy management of patients with early breast cancer. Biomedicine 1980;32:194-199.
[ISI][Medline]
[Order article via Infotrieve]
-
Chan DW, Beveridge RA, Bruzek DJ, Damron DJ, Bray KR, Gaur PK, et al. Monitoring breast cancer with CA 549. Clin Chem 1988;34:2000-2004.
[Abstract/Free Full Text]
-
Chatal JF, Chupin F, Ricolleau G, Tellier JL, Mevel Le A, Fumoleau P, et al. Use of serial carcinoembryonic antigen assays in detecting relapses in breast cancer involving high risk of metastases. Eur J Cancer 1981;17:233-238.
-
Chu TM, Nemoto T. Evaluation of carcinoembryonic antigen in human mammary carcinoma. J Natl Cancer Inst 1973;51:1119-1122.
-
Falkson HC, Falkson G, Portugal A, van der Watt JJ, Schoeman HS. Carcinoembryonic antigen as a marker in patients with breast cancer receiving postsurgical adjuvant chemotherapy. Cancer 1982;49:1859-1865.
[ISI][Medline]
[Order article via Infotrieve]
-
Haagensen DE, Kister SJ, Vandevoorde JP, Gates JB, Smart EK, Hansen HJ, Wells SA. Evaluation of carcinoembryonic antigen as a plasma monitor for human breast carcinoma. Cancer 1978;42:1512-1519.
[ISI][Medline]
[Order article via Infotrieve]
-
van der Linden JC, Baak JPA, Postma T, Lindeman J, Meyer CJLM. Monitoring serum CEA in women with primary breast tumours positive for oestrogen receptor and with spread to lymph nodes. J Clin Pathol 1985;38:1229-1234.
[Abstract/Free Full Text]
-
Merimsky O, Hareuveni M, Inbar M, Horev J, Keydar I, Kovner F, Chaitchik S. Increasing serum levels of mucin like carcinoma-associated antigen and mucinous antigen H23 in clinically disease-free breast cancer patients. Diagn Oncol 1993;3:61-66.
-
Nemoto T, Constantine R, Chu TM. Human tissue polypeptide
antigen in breast cancer. J Natl Cancer Inst 1979;63 1347–50..
-
Nicolini A, Carpi A, Di Marco G, Giuliani L, Giordan R, Palla S. A rational postoperative follow-up with carcinoembryonic antigen, tissue polypeptide antigen, and urinary hydroxyproline in breast cancer patients. Cancer 1989;63:2037-2046.
[ISI][Medline]
[Order article via Infotrieve]
-
Theriault RL, Hortobagyi GN, Fritsche HA, Frye D, Martinez R, Buzdar AU. The role of serum CEA as a prognostic indicator in stage II and III breast cancer patients treated with adjuvant chemotherapy. Cancer 1989;63:828-835.
[ISI][Medline]
[Order article via Infotrieve]
-
van Dalen A, Barak V, Cremaschi A, Gion M, Molina R, Namer M, et al. The prognostic significance of increasing marker levels in metastatic breast cancer patients with clinically complete remission, partial remission or stable disease. Int J Biol Markers 1998;13:10-15.
[ISI][Medline]
[Order article via Infotrieve]
-
Deprés-Brummer P, Itzhaki M, Bakker PJM, Hoek FJ, Veenhof KHN, de Wit R. The usefulness of CA15.3, mucin-like carcinoma-associated antigen and carcinoembryonic antigen in determining the clinical course in patients with metastatic breast cancer. J Cancer Res Clin Oncol 1995;121:419-422.
[ISI][Medline]
[Order article via Infotrieve]
-
Martoni A, Zamagni C, Bellanova B, Zanichelli L, Vecchi F, Cacciari E, et al. CEA, MCA, CA 15.3 and CA 549 and their combinations in expressing and monitoring metastatic breast cancer: a prospective comparative study. Eur J Cancer 1995;31A:1615-1621.
-
Sonoo H, Kurebayashi J. Serum tumor marker kinetics and the clinical course of patients with advanced breast cancer. Surg Today 1996;26:250-257.
[ISI][Medline]
[Order article via Infotrieve]
-
Murray A, Clinton O, Earl H, Price M, Moore A. Assessment of five serum marker assays in patients with advanced breast cancer treated with medroxyprogesterone acetate. Eur J Cancer 1995;31A:1605-1610.
-
Sölétormos G, Hyltoft Petersen P, Dombernowsky P. Progression criteria for cancer antigen 15.3 and carcinoembryonic antigen in metastatic breast cancer compared by computer simulation of marker data. Clin Chem 2000;46:939-949.
[Abstract/Free Full Text]
-
Fraser CG, Hyltoft Petersen P, Lytken Larsen M. Setting analytical goals for random analytical error in specific clinical monitoring situations. Clin Chem 1990;36:1625-1628.
[Abstract/Free Full Text]
The following articles in journals at HighWire Press have cited this article:
|
|
|
|
 
G. Soletormos, A. Semjonow, P. E.C. Sibley, R. Lamerz, P. H. Petersen, W. Albrecht, P. Bialk, M. Gion, F. Junker, H.-P. Schmid, et al.
Biological Variation of Total Prostate-Specific Antigen: A Survey of Published Estimates and Consequences for Clinical Practice
Clin. Chem.,
August 1, 2005;
51(8):
1342 - 1351.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
