Advances, Challenges, and Limitations in Serum-Proteome-Based
Matthias P. A. Ebert,†,* Murray Korc,|Peter Malfertheiner,§and Christoph Ro 1cken‡
Medical Department II, Klinikum rechts der Isar, Technical University of Munich, D-81675 Munich, Germany,
Institute of Pathology and Department of Gastroenterology and Hepatology, Otto-von-Guericke University,
D-39120 Magdeburg, Germany, and Departments of Medicine, and Pharmacology and Toxicology,
Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire 03756
Recent advances in medicine have dramatically reduced the incidence and mortality of many
cardiovascular, infectious, and certain neoplastic diseases; the overall mortality for most malignant
solid tumors remains high. The poor prognosis in these cancers is due, in part, to the absence of
adequate early screening tests, leading to delays in diagnosis. Three strategies have been applied to
fight cancer: analysis of the molecular mechanisms involved in its pathogenesis and progression,
improvement of early diagnosis, and the development of novel treatment strategies. There have been
major advances in our understanding of cancer biology and pathogenesis and in the development of
new (targeted) treatment modalities. However, insufficient progress has been made with respect to
improving the methods for the early diagnosis and screening of many cancers. Therefore, cancer is
often diagnosed at advanced stages, delaying timely treatment and leading to poor prognosis. Proteome
analysis has recently been used for the identification of biomarkers or biomarker patterns that may
allow for the early diagnosis of cancer. This tool is of special interest, since it allows for the identification
of tumor-derived secretory products in serum or other body fluids. In addition, it may be used to detect
reduced levels or loss of proteins in the serum of cancer patients that are present in noncancer
individuals. These changes in the serum proteome may result from cancer-specific metabolic or
immunological alterations, which are, at least partly, independent of tumor size or mass, thereby
fascilitating early discovery.
Keywords: tumor • serum • screening • diagnosis • proteome
Recent epidemiological data from Europe indicate that more
than 2 million Europeans will develop a malignant tumor, in
which gastrointestinal tumors in general and colorectal cancers
in particular are by far the most frequent cancers.1While there
are both increasing and decreasing incidences for various
cancers of the gastrointestinal tract, the overall mortality is high
and prognosis remains poor.1,2Most patients are diagnosed at
advanced stages with the presence of either a locally advanced
tumor and/or evidence of lymph node or distant metastases
that often do not allow for curative tumor resection. Cancer is
also often a disease of the elderly and serious co-morbidities
may further limit aggressive therapeutic options.3In these
cases, systemic therapy, such as chemotherapy or combined
chemoradiation, is necessary. Both may have serious side
effects, and the overall impact on patient survival and prognosis
is poor. Thus, to improve cancer prognosis, it has to be
diagnosed in its early stages.4This can be achieved by identify-
ing high-risk populations, enrolling them in screening and
surveillance programs, and using highly sensitive, specific, and
cost-effective disease markers, with the ultimate goal of either
treating cancer curatively or preventing its formation.5Thus,
the identification of markers of cancer development is of great
importance and would form the basis for efficient screening
programs. The requirements for these markers are the follow-
1. high sensitivity and specificity for the detection of cancer
at an early stage or detection of precursor lesions;
2. easy and noninvasive access to the site of biomarker
assessment, such as plasma or serum and other body fluids;
3. cheap, rapid, reproducible, and cost-effective determina-
tion of the biomarker or biomarker pattern; and
4. the capability to tail the treatment strategy.
If these requirements were met, the diagnosis of human
cancers would be greatly facilitated, and cancer prognosis
would improve. Unfortunately, most common tumor markers,
* Address correspondence to Matthias P. A. Ebert, MD, II. Medical
Department, Klinikum rechts der Isar, Technical University of Mu ¨nchen,
Ismaningerstr. 22, D-81675 Mu ¨nchen, Germany. Tel: +49-89-4140-2250.
Fax: +49-89-4140-4871. E-mail: email@example.com.
†Technical University of Munich.
|Dartmouth-Hitchcock Medical Center.
§Department of Gastroenterology and Hepatology, Otto-von-Guericke
‡Institute of Pathology, Otto-von-Guericke University.
that are mainly determined in the serum of cancer patients,
do not meet these requirements.6While the role of R-fetopro-
tein (AFP) for hepatocellular cancer and of prostate specific
antigen (PSA) for prostate cancer in clinical management of
these particular cancers is well-established, other cancers
cannot be identified by a tumor-specific serum marker.7,8For
example, gastric cancers are frequently diagnosed at an ad-
vanced stage, restricting the number of curatively treatable
patients.9By the time of diagnosis, most gastric cancers have
spread to the lymph nodes, which is the most significant
independent prognostic factor for this disease: the five-year-
survival rate of patients with lymph node metastases is 9.8 (
4% as compared with 58 ( 11% for patients without lymph
node metastases.10As in the case of many other cancers, the
early diagnosis of gastric cancer with currently available serum
tumor markers, such as CEA, Ca 72-4, or Ca 19-9, is impossible,
as they have a low sensitivity and specificity.11-13Stage I gastric
cancer is identified in less than 25% of the cases using any of
these serum markers. Thus, these serum markers are insuf-
ficient for early diagnosis or for the screening of gastric cancer.14
Alternatively, gastroscopic examination has been proposed as
a screening method for the early detection of gastric cancer.
However, no randomized trials evaluating a positive impact of
screening on mortality from gastric cancer have been re-
ported.15Gastroscopy is an expensive procedure with risks,
reaching beyond blood sampling. There may be some justifica-
tion for gastroscopic screening of some populations that are
at high risk for this malignancy. However, there is considerable
debate with respect to incidence rates that would justify such
screening, especially since gastroscopy does not meet require-
ments for a screening test, such as high convenience and
Molecular Diagnosis of Cancers. A multitude of studies have
addressed the molecular biology and pathogenesis of cancer.
Overall there seems to be a common background of genetic
and molecular alterations which underly cancer pathogenesis,
especially with respect to gastrointestinal cancers, where there
is an abundance of genomic mutations and a high frequency
of chromosomal instability.16Genetic and epigenetic changes
on the basis of microsatellite instability associated with im-
paired function of DNA repair genes provide another distinct
and different pathway for the development of cancers.17On
the basis of these findings, several groups have tried to use
specific molecular changes to identify patients with cancer or
precursor lesions. Analyses were performed using stool, serum,
blood, pancreatic juice, urine, and other biological samples.18-20
Generally, both RNA and DNA were employed for the analysis
of tumor-specific changes in biological fluids. These include
the analysis of tyrosinase mRNA in patients with suspected
malignant melanoma, thyroglobulin mRNA in patients with
suspected thyroid cancer, as well as other tumor-specific
changes.21,22Other studies addressed the utility of genomic DNA
alterations in the diagnosis of cancer patients, in which free
circulating DNA in serum or DNA from circulating cancer cells
in blood samples of patients with cancers was determined.23-25
Interestingly, the amount of free DNA in serum samples from
cancer patients is increased compared to individuals without
cancer.26K-ras mutations can be found in the blood of
pancreatic cancer patients and p16 and/or APC mutations in
peripheral blood or stool of colorectal cancer patients.23-25
Apart from the analysis of genetic alterations of the circulating
DNA in cancer patients, the presence of circulating viral DNA
or RNA has also been used for the identification of cancer
patients. Thus, circulating EBV DNA has been considered a
potential tumor marker for nasopharyngeal and gastric can-
cers.27,28Nonetheless, the analysis of genetic and molecular
alterations of genomic DNA in serum or blood of cancer
patients has been largely disappointing. Most of these studies
were performed with a small number of patients. The sensitivi-
ties and specificities achieved are, as yet, unsatisfactory. In
addition, not all cancers, even of the same origin, exhibit the
same molecular changes, increasing the risk for false-negative
Considerable interest was raised by the analysis of epigenetic
changes in biological samples from cancer patients, since these
changes are present frequently and occur early in the patho-
genesis of almost all cancers.29First results from the analysis
of gene methylation in epithelial cells present in either the
sputum of lung cancer patients, the stool, or peripheral blood
of colorectal cancer patients, have been published recently.30,31
However, again these studies need to be confirmed by large
studies with more patients and adequate controls before a final
interpretation of their efficacy can be established. Apart from
the detection of cancer-specific alterations of tumor-derived
RNA and DNA, serum levels of certain proteins involved in
tumor biology have also been assessed in cancer patients.
Serum levels of cathepsin B, E-cadherin, hepatocyte growth
factor, interleukins, and other cytokines and hormones have
been measured in the serum of cancer patients.32-36However,
while some markers are useful to assess patient prognosis after
a diagnosis was reached, sensitivity and specificity are too low
in order to fascilitate the detection of cancer in its early stages.
Serum Protein Profiling. Recent developments in the field
of proteome analysis have led to considerable advances in our
understanding of the changes and functional expression of
proteins in the process of cancer biology. 2D-PAGE analysis
has been the standard procedure for more than 30 years, which
has been combined with mass spectrometry (MALDI) for the
detection of aberrantly expressed proteins in tissue and serum
of cancer patients.37With the help of 2D-PAGE analysis,
nanomolar amounts of proteins are separated and identified
by mass spectrometry, especially in the molecular weight
ranging from 10 to 150 kDa. Approximately 1000-3000 proteins
can be separated on a pH range from 3 to 10, whereas using a
more narrow pH range, for example, 4-7, the resolution and
number of proteins can be increased. However, this method
exhibits some major limitations regarding its role in the
identification of potential biomarkers for cancer diagnosis: this
method is less well-suited for small, and either very basic or
acidic proteins, or hydrophobic proteins.38,39Furthermore, the
protein amounts required for reproducible identification of
serum proteins are rather high, and standardization of the
technique is difficult. In contrast, SELDI-TOF MS is a rather
new method which is especially valuable for the identification
of serum-derived biomarkers.40This method is based on
ProteinChip Arrays which carry various chromatographic prop-
erties, such as anion exchange, cation exchange, and hydro-
philic or hydrophobic surfaces.41For the analysis of serum, only
5-10 µL of serum sample is applied to these surfaces; after
washing off unbound material, the protein fingerprint can be
determined and visualized by time-of-flight mass spectrometry.
The advantages of this method are the low amount of sample
necessary for analysis, its speed, and high throughput capabil-
ity.40,41Many different groups have used this method and
related methods based on prefractionation of serum proteins
by beads and subsequent MALDI analysis for the identification
of biomarkers in serum, urine, pancreatic juice, and other
biological fluids.42-46In these studies, the identification of
cancers based on these biomarkers or biomarker patterns was
possible and beyond the sensitivity and specificity of conven-
tional serum markers. Interestingly, many of the peptides and
proteins that were identified in the course of SELDI analysis
have not previously been linked to the biology of human
cancers, and thus, this method also provides further insight
into the changes leading to or underlying cancer development
and progression. Various groups have confirmed the high
sensitivity and specificity of this method for the detection of
various cancers, including bladder, prostate, ovarian, pancre-
atic, and breast cancer.42-45We also used SELDI analysis based
on ProteinChip Arrays to screen for biomarkers or biomarker
patterns for the identification of gastric cancer patients.46
Interestingly, we found profound changes in the serum protein
profile in gastric cancer patients versus patients with dyspepsia
in which cancer was ruled out by endoscopy (Figure 1).
However, none of the single masses proved to be able to
separate all cancer patients from noncancer individuals. Using
the decision tree analysis, we generated a classifier ensemble
with 28 out of 71 masses to generate 50 decision trees. This
classifier ensemble was highly effective in the differentiation
of cancer patients from noncancer individuals. Thus, while all
cancers in the training set were correctly classified (sensitivity
100%), even 8 of 9 stage I cancers of an independent test set
were correctly classified as cancer patients.46Specificity was
high as well, in that all but one individual without cancer of a
further independent test set were correctly classified as non-
cancer individuals. Interestingly, the decision trees were gener-
ated from masses that were identified both in the serum of
cancer individuals and noncancer individuals. Thus, the high
performance of the classifier ensemble was based not only on
the identification of tumor-derived or tumor-specific masses
in cancer patients but also on the decreased levels or loss of
certain proteins in the serum of cancer patients which were
present in noncancer individuals. Several other groups have
also reported the down-regulation of certain proteins in
different kinds of cancer both in the tumor itself using MALDI
MS imaging or in the serum using SELDI-TOF MS.47-50Re-
cently, magnet bead-assisted prefractionation of serum proteins
combined with proteome analysis by MALDI has also been
developed and used for the identification of biomarkers and
biomarker patterns in various diseases, such as lymphatic
leukaemia, brain tumors, and inflammatory diseases.51-53While
reproducibility has been demonstrated to be good in a study
conducted by Zhang et al.,53further studies are required to
define the role of this approach in clinical proteomics.
The Tumor-Host Interface and Cancer Diagnosis. Con-
ventional serum tumor markers are based on the measurement
of tumor-derived or tumor-specific secretory peptides or
proteins in the serum of cancer patients. Thus, the detection
of circulating levels of AFP has proven to be of adequate efficacy
for the identification of most patients with hepatocellular
cancer, at least in advanced stages.7In addition, other con-
ventional serum markers such as carcinoembryonic antigen
(CEA) or PSA are regarded to be tumor-specific proteins that
are produced and secreted by the tumor and reach the
circulation in which elevated levels are regarded as diagnostic
criteria for cancer detection.8In this respect, these proteins can
be regarded as part of the serum profile of cancer patients in
that the presence of these proteins in the serum of cancer
patients contributes to the tumor-specific protein profile of
cancer patients. However, PSA may also help to illustrate one
major problem of positive serum protein profiling. PSA is only
synthesized by prostatic columnar epithelium and is secreted
into the seminal plasma.54Even in healthy individuals, a very
small proportion (1 in 100-1000 molecules) of PSA reaches the
circulation through cell and tissue leakage, where it is rapidly
diluted in 5 L blood volume. In healthy males, serum PSA
averages 1 µg/L, while, in seminal plasma, it is in the range of
g/L.54Under disease conditions, cell and tissue leakage in-
creases, as does the serum level of PSA. The sensitivity and
specificity of serum PSA levels depend on the amount of protein
synthesized by the individual tumor cell, the tumor cell mass,
and co-variables, such as inflammation and tissue manipula-
tion. Thus, a putative, tumor-specific biomarker might indeed
reach the serum by cell and tissue leakage. However, several,
not necessarily tumor-specific variables, influence sensitivity
and specificity of this biomarker. Therefore, to be informative,
the serum level has to exceed a certain cutoff value.54
Sensitivity can be increased by adding up several biomarkers
to a biomarker pattern (Figure 2). When SELDI analysis was
used, a wide range of further proteins were identified which
ultimately may result in a more complex “positive serum
profile”. Recent studies published by Rosty et al. and other
groups have identified biomarkers such as the pancreatitis-
associated protein (HIP/PAP-I) or defensins which are present
in cancer patients.43Overall, the combined analysis of these
serum markers and the generation of biomarker patterns based
on tumor-specific or tumor-derived serum proteins have
increased the diagnostic yield substantially (Figure 2).
Apart from tumor- and tissue-specific gene products, such
as AFP, CEA, PSA, or HIP/PAP-I, entering the circulation
through cell leakage, tissue leakage, and necrosis, serum protein
profiling may also detect signatures in the serum of post-
translational protein- or peptide-modifications taking place at
the tumor-host interface (Figure 3). These might be due to
the development of a desmoplastic stroma, neoangiogenesis,
and the inflammatory response to tumor invasion. Many
Figure 1. Overlay of representative spectra of healthy individuals (red) and patients with gastric cancer (blue). Note reduced peak
height of several masses, indicating reduced serum levels or even loss of certain masses in cancer patients (arrows).
enzymes are differentially expressed in malignant tumors, some
of which have unique functions in extracellular matrix and cell
homeostasis, are proteolytically active in the tumor-host
interface, and promote tumor growth and spread.55Among
many other enzymes, cathepsins were found to be upregulated
in tumor cells of various malignancies, and increased serum
levels were found in cancer patients, probably as a result of
cell and tissue leakage, cell death, and aberrant secretion.32
These enzymes modify proteins and peptides generating puta-
tive biomarkers of the disease. The tumor-host interface is
unique for a malignant tumor. Thus, it is very reasonable to
assume that serum-proteomic profiling can separate invasive
cancer from inflammation as shown by several studies.42,56
Enzymatic activity at the tumor-host interface, as well as
metabolic and immunological changes in cancer patients, may
lead to substantial changes of the serum proteome, including
various diagnostically relevant post-translational modifica-
tions.57These changes may not only generate diagnostically
relevant tumor-specific or tumor-derived proteins in the serum
of cancer patients but also lead to the reduction or loss of
certain peptide or protein species naturally present in healthy,
noncancer individuals (Figure 2). This “loss” of certain peptides
or proteins in the serum of cancer patients may result from a
change in protein synthesis, an altered protein metabolism, or
a tumor-specific post-translational modification and may
further increase the diagnostic efficacy of biomarker patterns
for the diagnosis of cancer patients (Figure 2). In our study, in
which we analyzed serum of gastric cancer patients, the single
best mass for the differentiation of gastric cancer from non-
cancer individuals was thrombin light chain a, a proteolytic
fragment of prothrombin, which was reduced or undetectable
in cancer patients.58This indicates not only that the loss of
certain proteins may have substantial diagnostic impact but
also that the coagulation system undergoes profound changes
Figure 2. Comparison of the diagnostic accuracy of various approaches to serum-based cancer diagnosis. The diagnostic accuracy of
conventional tumor markers (A) is low, and even the sensitivity of the combined analysis of several tumor markers is often below
60-70%. Various molecular markers (B) as well as single biomarkers (C) identified by proteome analysis only detect a small subset of
cancer patients, since not all cancers express these proteins or exhibit these alterations in the cancer cells. The generation of biomarker
patterns, e.g., through decision tree analysis, markedly increases sensitivity (D). Apart from the detection of proteins or peptides that
are secreted by the tumor (red boxes), the integration of serum proteins that are decreased in cancer individuals but are present in
noncancer individuals (blue boxes) in a more complex decision tree may improve the identification of cancers (E).
in cancer patients. These observations support our hypothesis
that metabolic and immunological changes can be found in
every patient with a malignant tumor and, therefore, could be
used as a diagnostic tool.48-50,58SELDI-TOF MS and related
MALDI technologies have the capacity to detect and visualize
both the positive and negative serum profile of cancer patients
and noncancer individuals and thereby allow the identification
of cancer patients with a sensitivity and specificity that cannot
be reached by conventional or single tumor markers (Figure
2). The importance of the negative serum profile is underscored
by the fact that these changes are independent of the size of
the primary tumor, the presence of lymph node or distant
metastasis, or the overall tumor stage. In our analysis, even 8
of 9 stage I cancers were correctly classified, indicating that
this approach may not only improve cancer detection but also
have a special impact on the identification of early cancers
which could be treated with a curative intent. If the detection
of early gastric cancers by SELDI is confirmed in larger studies,
endoscopic screening may be tailored to high-risk individuals
with serum protein profiles indicating early cancers, which
could then be treated curatively. While large endoscopic
screening programs are not cost-effective, the tailored approach
to individuals with pathological serum protein profiles may well
improve diagnostic accuarcy and overall sensitivity.15
In prostate cancers, several studies have also confirmed the
high specificity and sensitivity for the discrimination of prostate
cancer patients from noncancer patients.59In a recent study,
Ornstein et al. also demonstrated that these serum-proteome
patterns allow the discrimination of men with elevated PSA due
to benign processes from men with prostate cancer.60In the
conclusion of their study, they postulate that these serum
profiles may allow the reduction of unnecessary prostate
biopsies without compromising the detection of curable pros-
tate cancers, an approach which would also tailor biopsies to
high-risk individuals identified by SELDI analysis.
Limitations of Serum-Proteome-Based Cancer Diagnosis.
Although SELDI analysis has identified a number of potential
biomarkers and biomarker patterns that allow the identification
of cancer patients with a very high sensitivity and specificity,
a number of questions have been raised regarding the repro-
ducibility and specifity of this approach.61,62Thus, this method
has been regarded not to be sensitive enough to detect low-
abundance proteins which would be important for early cancer
diagnosis. Furthermore, the proteins which are detected in the
serum using this method are often not regarded as tumor-
specific, such as transthyretin, haptoglobin, or amyloid A
protein.63The greatest concern, however, is reproducibility
within each group and among various groups using this
method. Thus, often, biomarkers have not been (re)-identified
in experiments using the same study population and applying
different bioinformatic approaches.45However, recently, two
independent studies using sera from patients with hepatocel-
lular cancers identified a 8900 kDa mass in the cancer sera in
both studies, which was finally identified as the C-terminal part
Figure 3. The role of serum tumor profiling for the diagnosis of cancer. Adenocarcinomas often synthesize and secrete mucous
substances as well as many other peptides and proteins, which may eventually reach the serum, where they can be detected by serum
profiling (Tumor marker). The hallmark of cancer, i.e., invasive growth, is characterized by the development of a desmoplastic stroma
and a local inflammatory response (Tumor metabolism). High local proteolytic activity and the specific metabolic and immunologic
changes due to invasive growth change the serum proteome in a complex manner often with reduction of specific peptides and proteins
in the serum. A combination of serum profiling and detection of tumor metabolic changes bear the capability to significantly improve
early detection of cancer, reaching beyond conventional serum profiling.
of the V10 fragment of vitronectin.64,65Furthermore, Zhang et
al. identified biomarkers specific for early ovarian cancer using
sera which were drawn and analyzed by SELDI at two centers
independently, and data were cross-validated to identify
potential biomarkers.66Furthermore, in a recent multiinstitu-
tional approach, interlaboratory calibration and standardization
of the SELDI assay platform was confirmed.67Nonetheless,
despite these recent reports, reproducibility of biomarker
identification has been an issue of debate and further ap-
proaches are necessary to solve this problem.45
Regarding the issue of tumor specificity, Fung et al. recently
reevaluated the role of previously identified biomarkers, that
is, transthyretin and inter-alpha trypsin inhibitor heavy chain
4 (ITIH4), in sera of patients with ovarian, breast, prostate, and
colon cancer.57Most of these proteins, including albumin,
transthyretin, lipoproteins, c-reactive proteins, and others, are
synthesized in the liver and have been regarded as noncancer
specific alterations or epiphenomena of tumors which result
from a cascade of inflammatory signals.68However, in their
study, they found a high degree of post-translational modifica-
tions, including proteolytic truncation, cysteinylation, and
glutathionylation in these cancer sera. These post-translational
modifications were identified by SELDI-TOF MS and allowed
the differentiation of cancer sera from noncancer sera and also
the differential diagnosis of tumor type.57Thus, these post-
translational modifications of serum proteins in cancer patients
allowed cancer-specific diagnosis which would not have been
identified by the use of a conventional ELISA technique.
Although these serum proteins are synthesized mainly by the
liver and not by the tumor itself and, thus, are regarded as
nonspecific systemic reactions, the modifications of these
proteins may seem suitable for cancer detection and, thus,
should also be considered as biomarkers for cancers. Inasmuch
as this host response, which has been termed host response
protein amplification cascade by Fung et al.,57includes both
the induction and down-regulation of circulating proteins as
a consequence of tumor development early in the process of
cancer pathogenesis, the combined detection of their tumor-
specific post-translational modifications in conjunction with
other tumor-secreted biomarkers may still allow early diagnosis
of cancer. Their data also support our hypothesis that the
metabolic activity in cancer patients and at the tumor-host
interface may be associated with cancer and cancer-subtype-
specific, diagnostically relevant post-translational modifications
To further solve the remaining other problems of reproduc-
ibility, sensitivity, and specificity, a number of other actions
should be taken as recommended by Diamandis and our group
in recent publications.61,62,69This includes the importance of
identifying the biomarker, the inclusion of internal controls in
the SELDI analysis, and the careful control of the study
population with regard to disease state, sampling conditions,
sampling storage, and other clinical features.61,62,69
Conclusions and Outlook
Most cancers are diagnosed in advanced stages when cura-
tive treatment is impossible and prognosis poor. Conventional
serum markers are largely unsatisfactory for the early detection
of cancer or the screening of high-risk individuals. Even
molecular markers, based on the detection of mRNA levels of
certain factors that have been implicated in tumor biology or
genetic or epigenetic changes of DNA in serum or peripheral
blood of cancer patients, have not led to improvements in the
early diagnosis of human cancers. Mostly, the available studies
were performed only with few patients, and sensitivity and
specificity were limited in these studies. The analysis of the
serum proteome has led to the identification of a number of
proteins, peptides, and disease-specific post-translational modi-
fications that may be used as biomarkers or biomarker patterns
for cancer detection. Serious questions of reproducibility,
sensitivity, and specificity need, however, be addressed before
this technique may be introduced in the clinical management
of cancer patients. Nonetheless, since classical serum tumor
markers have generally failed to identify cancers in their early
stages, the detection of metabolic and immunologic changes
in the serum proteome of cancer patients by SELDI and MALDI
analyses may be useful to improve our understanding of cancer
biology and may provide a very useful tool for the early
diagnosis of human cancers (Figure 3). As soon as these serum
protein profiles or biomarker patterns have been confirmed
in large studies involving various centers throughout the world,
it is conceivable that serum protein profiling may allow the
easy, cheap, noninvasive, reproducible, and rapid identification
of cancer patients with a high sensitivity and specificity. While
most current screening strategies are not cost-effective and
often include invasive procedures to obtain histology, serum-
proteome-based cancer diagnosis may allow the identification
of high-risk individuals that should undergo more invasive
procedures to diagnose and even curatively treat neoplastic or
Acknowledgment. M. Ebert is supported by the Heisen-
berg-Programme of the DFG (Eb 187/5-1) and by a grant from
the Land Sachsen-Anhalt.
(1) Boyle, P.; Autier, P.; Bartelink, H.; Baselga, J.; Boffetta, P.; Burn,
J.; Burns, H. J.; Christensen, L.; Denis, L.; Dicato, M.; Diehl, V.;
Doll, R.; Franceschi, S.; Gillis, C. R.; Gray, N.; Griciute, L.;
Hackshaw, A.; Kasler, M.; Kogevinas, M.; Kvinnsland, S.; La
Vecchia, C.; Levi, F.; McVie, J. G.; Maisonneuve, P.; Martin-
Moreno, J. M.; Bishop, J. N.; Oleari, F.; Perrin, P.; Quinn, M.;
Richards, M.; Ringborg, U.; Scully, C.; Siracka, E.; Storm, H.;
Tubiana, M.; Tursz, T.; Veronesi, U.; Wald, N.; Weber, W.; Zaridze,
D. G.; Zatonski, W.; zur Hausen, H. Ann. Oncol. 2003, 14, 973-
(2) Boyle, P.; d’Onofrio, A.; Maisonneuve, P.; Severi, G.; Robertson,
C.; Tubiana, M.; Veronesi, U. Ann. Oncol. 2003, 14, 1312-1325.
(3) Gridelli, C.; Aapro, M.; Ardizzoni, A.; Balducci, L.; De Marinis, F.;
Kelly, K.; Le Chevalier, T.; Manegold, C.; Perrone, F.; Rosell, R.;
Shepherd, F.; De Petris, L.; Di Mario, M.; Langer, C. J. Clin. Oncol.
2005, 23, 3125-3137.
(4) Bonelli, L. Tech. Coloproctol. 2004, 8, s267-272.
(5) Crawford, E. D. Lancet 2005, 365, 1447-1449.
(6) Bast, R. C., Jr.; Ravdin, P.; Hayes, D. F.; Bates, S.; Fritsche, H., Jr.;
Jessup, J. M.; Kemeny, N.; Locker, G. Y.;, Bennel, R. G.; Somerfield,
M. R. J. Clin. Oncol. 2001, 19, 1865-1878.
(7) Di Bisceglie, A. Gastroenterology 2004, 127, S104-107.
(8) Hernandez, J.; Thompson, I. Cancer 2004, 101, 894-904.
(9) Ebert, M. P. A.; Malfertheiner, P. Aliment. Pharmacol. Ther. 2002,
(10) Yokota, T.; Ishiyama, S.; Saito, T.; Teshima, S.; Narushima, Y.;
Murata, K.; Iwamoto, K.; Yashima, R.; Yamauchi, H.; Kikuchi, S.
Scand. J. Gastroenterol. 2004, 39, 380-384.
(11) Marrelli, D.; Roviello, F.; de Stefano, A.; Farnetani, M.; Garosi, L.;
Messano, A.; Pinto, E. Oncology 1999, 57, 55-62.
(12) Nakajiima, K.; Ochiai, T.; Suzuki, T.; Shimada, H.; Hayashi, H.;
Yasumoto, A.; Takeda, A.; Hishikawa, E.; Isono, K. Tumor Biol.
1998, 19, 464-469.
(13) Istigami, S.; Natsugoe, S.; Hokita, S.; Che, X.; Tokuda, K.; Nakajo,
A.; Iwashige, H.; Tokushige, M.; Watanabe, T.; Takao, S.; Aikou,
T. J. Clin. Gastroenterol. 2001, 32, 41-44.
(14) Lieberman, D.; Fennerty, M.. B.; Morris, C. D.; Holub, J.; Eisen,
G.; Sonnenberg, A. Gastroenterology 2004, 127, 1067-1075.
(15) Riecken, B.; Pfeiffer, R.; Ma, J. L.; Jin, M. L.; Li, J. Y.; Liu, W. D.;
Zhang, L.; Chang, Y. S.; Gail, M. H.; You, W. C. Prev. Med. 2002,
(16) Draviam, V. M.; Xie, S.; Sorger, P. K. Curr. Opin. Genet. Dev. 2004,
(17) Feinberg, A. P. Semin. Cancer Biol. 2004, 14, 427-432.
(18) von Knobloch, R.; Brandt, H.; Schrader, A. J.; Heidenreich, A.;
Hofmann, R. Clin. Cancer Res. 2004, 10, 988-993.
(19) Kanaoka, S.; Yoshida, K.; Miura, N.; Sugimura, H.; Kajimura, M.
Gastroenterology 2004, 127, 422-427.
(20) Melissourgos, N.; Kastrinakis, N. G.; Davilas, I.; Foukas, P.;
Farmakis, A.; Lykourinas, M. Urology 2003, 62, 362-367.
(21) Stoitchkov, K.; Letellier, S.; Garnier, J. P.; Toneva, M.; Naumova,
E.; Peytcheva, E.; Tzankov, N.; Bousquet, B.; Morel, P.; Le Bricon,
T. Clin. Chim. Acta 2001, 306, 133-138.
(22) Kaufmann, S.; Schmutzler, C.; Schomburg, L.; Korber, C.; Luster,
M.; Rendl, J.; Reiners, C.; Kohrle, J. Endocr. Regul. 2004, 38, 41-
(23) Kanyama, Y.; Hibi, K.; Nakayama, H.; Kodera, Y.; Ito, K.; Akiyama,
S. Cancer Sci. 2003, 94, 418-420.
(24) Wang, J. Y.; Hsieh, J. S.; Chang, M. Y.; Huang, T. J.; Chen, F. M.;
Cheng, T. L.; Alexandersen, K.; Huang, Y. S.; Tzou, W. S.; Lin, S.
R. World J. Surg. 2004, 28, 721-726.
(25) Lauschke, H.; Caspari, R.; Friedl, W.; Schwarz, B.; Mathiak, M.;
Propping, P.; Hirner, A. Cancer Detect. Prev. 2001, 25, 55-61.
(26) Gautschi, O.; Bigosch, C.; Huegli, B.; Jermann, M.; Marx, A.;
Chasse, E.; Ratschiller, D.; Weder, W.; Joerger, M.; Betticher, D.
C.; Stahel, R. A.; Ziegler, A. J. Clin. Oncol. 2004, 22, 4157-4164.
(27) Chan, K. C.; Lo, Y. M. Semin. Cancer Biol. 2002, 12, 489-496.
(28) Lo, Y. M.; Chan, W. Y.; Ng, E. K.; Chan, L. Y.; Lai, P. B.; Tam, J. S.;
Chung, S. C. Clin. Cancer Res. 2001, 7, 1856-1859.
(29) Lofton-Day, C.; Lesche, R. Dig. Dis. 2003, 21, 299-308.
(30) Wang, Y. C.; Lu, Y. P.; Tseng, R. C.; Lin, R. K.; Chang, J. W.; Chen,
J. T.; Shih, C. M.; Chen, C. Y. J. Clin. Invest. 2003, 111, 887-895.
(31) Leung, W. K.; To, K. F.; Man, E. P.; Chan, M. W.; Bai, A. H.; Hui,
A. J.; Chan, F. K.; Lee, J. F.; Sung, J. J. Clin. Chem. 2004, 50, 2179-
(32) Ebert, M. P.; Kruger, S.; Fogeron, M. L.; Lamer, S.; Chen, J.; Pross,
M.; Schulz, H. U.; Lage, H.; Heim, S.; Roessner, A.; Malfertheiner,
P.; Rocken, C. Proteomics 2005, 5, 1693-1704.
(33) Juhasz, M.; Ebert, M. P.; Schulz, H. U.; Rocken, C.; Molnar, B.;
Tulassay, Z.; Malfertheiner, P. Scand. J. Gastroenterol. 2003, 38,
(34) Maruo, Y.; Gochi, A.; Kaihara, A.; Shimamura, H.; Yamada, T.;
Tanaka, N. Int. J. Cancer 2002, 100, 486-490.
(35) Karayiannakis, A. J.; Syrigos, K. N.; Polychronidis, A.; Zbar, A.;
Kouraklis, G.; Simopoulos, C. Ann. Surg. 2002, 236, 37-42.
(36) Ohta, M.; Konno, H.; Tanaka, T.; Baba, M.; Kamiya, K.; Syouji, T.
Cancer Lett. 2003, 192, 215-225.
(37) Gorg, A.; Weiss, W.; Dunn, M. J. Proteomics 2004, 4, 3665-3685.
(38) Aldred, S.; Grant, M. M.; Griffiths, H. R. Clin. Biochem. 2004, 37,
(39) Wright, M. E.; Han, D. K.; Aebersold, R. Mol. Cell. Proteomics 2005,
(40) Wulfkuhle, J. D.; Liotta, L. A.; Petricoin, E. F. Nat. Rev. Cancer
2003, 3, 267-275.
(41) Issaq, H. J.; Veenstra, T. D.; Conrads, T. P.; Felschow, D. Biochem.
Biophys. Res. Commun. 2002, 292, 587-592.
(42) Vlahou, A.; Schellhammer, P. F.; Mandrinos, S.; Patel, K.; Kondylis,
F. I.; Gong, L.; Nasim, S.; Wright, G. L. Am. J. Pathol. 2001, 158,
(43) Rosty, C.; Christa, L.; Kuzdzal, S.; Baldwin, W. M.; Zahurak, M.
L.; Carnot, F.; Chan, D. W.; Canto, M.; Lillemoe, K. D.; Cameron,
J. L.; Yeo, C. J.; Hruban, R. H.; Goggings, M. Cancer Res. 2002,
(44) Petricoin, E. F.; Ardekani, A. M.; Hitt, B. A.; Levine, P. J.; Fusaro,
V. A.; Steinberg, S. M.; Mills, G. B.; Simone, C.; Fishman, D. A.;
Kohn, E. C.; Liotta, L. A. Lancet 2002, 359, 572-577.
(45) Adam, B. L.; Qu, Y.; Davis, J. W.; Ward, M. D.; Clements, M. A.;
Cazares, L. H.; Semmes, O. J.; Schellhammer, P. F.; Yasui, Y.; Feng,
Z.; Wright, GL. Cancer Res. 2002, 62, 3609-3614.
(46) Ebert, M. P.; Meuer, J.; Wiemer, J. C.; Schulz, H. U.; Reymond,
M. A.; Traugott, U.; Malfertheiner, P.; Rocken, C. J. Proteome Res.
2004, 3, 1261-1266.
(47) Mobley, J. A.; Lam, Y. W.; Lau, K. M.; Pais, V. M.; L’Esperance, J.
O.; Steadman, B.; Fuster, L. M.; Blute, R. D.; Taplin, M. E.; Ho, S.
M. J. Urol. 2004, 172, 331-337.
(48) Lam, Y. W.; Mobley, J. A.; Evans, J. E.; Carmody, J. F.; Ho, S. M.
Proteomics 2005, 5, 2927-2938.
(49) Yanagisawa, K.; Shyr, Y.; Xu, B. J.; Massion, P. P.; Larsen, P. H.;
White, B. C.; Roberts, J. R.; Edgerton, M.; Gonzalez, A.; Nadaf, S.;
Moore, J. H.; Caprioli, R. M.; Carbone, D. P. Lancet 2003, 362,
(50) Schwartz, S. A.; Weil, R. J.; Johnson, M. D.; Toms, S. A.; Caprioli,
R. M. Clin. Cancer Res. 2004, 10, 981-987.
(51) Pusch, W.; Flocco, M. T.; Leung, S. M.; Thiele, H.; Kostrzewa, M.
Pharmacogenomics 2003, 4, 463-476.
(52) Villanueva, J.; Philip, J.; Entenberg, D.; Chaparro, C. A.; Tanwar,
M. K.; Holland, E. C.; Tempst, P. Anal. Chem. 2004, 76, 1560-
(53) Zhang, X.; Leung, S. M.; Morris, C. R.; Shigenaga, M. K. J. Biomol.
Tech. 2004, 15, 167-175.
(54) Schmid, H. P.; Riesen, W.; Prikler, L. Crit. Rev. Oncol. Hematol.
2004, 50, 71-78.
(55) Koblinski, J. E.; Ahram, M.; Sloane, B. F. Clin. Chim. Acta 2000,
(56) Koopmann, J.; Zhang, Z.; White, N.; Rosenzweig, J.; Fedarko, N.;
Jagannath, S.; Canto, M. I.; Yeo, C. J.; Chan, DW.; Goggins, M.
Clin. Cancer Res. 2004, 10, 860-868.
(57) Fung, E. T.; Yip, T. T.; Lomas, L.; Wang, Z.; Yip, C.; Mang, X. Y.;
Lin, S.; Zhang, F.; Zhang, Z.; Chan, D. W.; Weinberger, S. R. Int.
J. Cancer 2005, 115, 783-789.
(58) Ebert, M. P.; Lamer, S.; Meuer, J.; Malfertheiner, P.; Reymond,
M.; Buschmann, T.; Rocken, C.; Seibert, V. J. Proteome Res. 2005,
(59) Qu, Y.; Adam, B. L.; Yasui, Y.; Ward, M. D.; Cazares, L. H.;
Schellhammer, P. F.; Feng, Z.; Semmes, O. J.; Wright, G. L. Clin.
Chem. 2002, 48, 1835-1843.
(60) Ornstein, D. K.; Rayford, W.; Fusaro, V. A.; Conrads, T. P.; Ross,
S. J.; Hitt, B. A.; Wiggins, W. W.; Veenstra, T. D.; Liotta, L. A.;
Petricoin, E. F. J. Urol. 2004, 172, 1302-1305.
(61) Jacobs, I. J.; Menon, U. Mol. Cell. Proteomics 2004, 3, 355-366.
(62) Diamandis, E. P. Mol. Cell. Proteomics 2004, 3, 367-378.
(63) Zhang, Z.; Bast, R. C.; Yu, Y.; Li, J.; Sokoll, L. J.; Rai, A. J.;
Rosenzweig, J. M.; Cameron, B.; Wang, Y. Y.; Mang, X. Y.; Berchuk,
A.; Van Haaften-Day, C. Cancer Res. 2004, 64, 5882-5890.
(64) Poon, T. C. W.; Yip, T. T.; Chan, A. T. C.; Yip, C.; Yip, V.; Mok, T.
S. K.; Lee, C. C. Y.; Leung, T. W. T.; Ho, S. K. W.; Johnson, P. J.
Clin. Chem. 2003, 49, 752-760.
(65) Paradis, V.; Degos, F.; Dargere, D.; Pham, N.; Belghiti, J.; Degott,
C.; Janeau, J. L.; Bezeaud, a.; Delforge, D.; Cubizolles, M.;
Laurendeau, I.; Bedossa, P. Hepatology 2005, 41, 40-47.
(66) Zhang, Z.; Bast, R. C.; Yu, Y.; Li, J.; Sokoll, L. J.; Rai, A. J.;
Rosenzweig, J. M.; Cameron, B.; Wang, Y. Y.; Meng, X. Y.; Berchuk,
A.; van Haaften-Day, C.; Hacker, N. F.; de Bruijn, H. W. A.; van
der Zee, A. G. J.; Jacobs, I. J.; Fung, E. T.; Chan, D. W. Cancer Res.
2004, 64, 5882-5890.
(67) Semmes, O. J.; Feng, Z.; Adam, B. L.; Banez, L. L.; Bigbee, W. L.;
Campos, D.; Cazares, L. H.; Chan, D. W.; Grizzle, W. E.; Izbicka,
E.; Kagan, J.; Malik, G.; McLerran, D.; Moul, J. W.; Partin, A.;
Prasanna, P.; Rosenzweig, J.; Sokoll, L. J.; Srivastava, S.; Srivastava,
S.; Thompson, I.; Welsh, M. J.; White, N.; Winget, M.; Yasui, Y.;
Zhang, Z.; Zhu, L. Clin. Chem. 2005, 1, 102-112.
(68) Sheth, K.; Bankey, P. Curr. Opin. Crit. Care 2001, 7, 99-104.
(69) Seibert, V.; Ebert, M. P. A.; Buschmann, T. Briefings Funct.
Genomics Proteomics 2005, 4, 1-11.