Preoperative Plasma Plasminogen Activator Inhibitor Type-1
and Serum C-Reactive Protein Levels in Patients With
Hans Jørgen Nielsen, MD, DMSc, Ib Jarle Christensen, MSc, Steen Sørensen, MD, DMSc,
Flemming Moesgaard, MD, DMSc, Nils Bru ¨nner, MD, DMSc, and
The RANX05 Colorectal Cancer Study Group
Background: Preoperative plasma plasminogen activator inhibitor-1 (PAI-1) is a prognostic
variable in patients with colorectal cancer. It has been suggested, however, that plasma PAI-1 is a
nonspecific prognostic parameter similar to the acute-phase reactant C-reactive protein (CRP). In the
present study we analyzed the association between plasma PAI-1 and serum CRP in patients
scheduled for elective resection of colorectal cancer. In addition, the prognostic value of PAI-1 and
CRP was studied in this patient cohort.
Methods: PAI-1 and CRP were analyzed in citrated plasma and serum, respectively, obtained
preoperatively from 594 patients. Patients who required preoperative blood transfusion received
SAGM blood, in which soluble PAI-1 is not present. None of the patients received pre- or
postoperative adjuvant chemotherapy, and all were followed in the outpatient clinic for at least 5
years or until death. The association of PAI-1 and CRP, respectively, with survival was tested using
the median value of PAI-1 and the upper normal limit for CRP. Analyses were performed by
inclusion of all patients, and in the subgroup of patients, who underwent curative resection.
Results: The median follow-up period was 6.8 (5.4–7.9) years. The median value of plasma
PAI-1 was 35.8 ng/ml, and values greater than 94 nmol/L identified patients with increased CRP
levels. Comparison of the molecules showed that PAI-1 was weakly correlated with CRP (r ? .26;
P ?.0001). Both molecules showed a Dukes independent distribution. In univariate survival
analyses high levels of PAI-1 were found associated with poor prognosis and low levels with good
prognosis (P ? .02, HR: 1.3). Similarly, high levels of CRP were found associated with poor
prognosis and low levels with good prognosis (P ?.0001, HR: 1.9). In a multivariate statistical
analysis including Dukes classification, gender, age, tumor location, perioperative blood transfusion,
PAI-1 and CRP, plasma PAI-1 was a dependent prognostic variable, while serum CRP (P ?.0001;
HR: 1.4; 95% CI: 1.3–1.5) was found to be a Dukes independent prognostic variable. Similar
analyses, excluding patients with Dukes’ D disease showed serum CRP to be an independent
prognostic variable (P ?.0001; HR: 1.3: 95% CI: 1.2–1.5).
Conclusions: This study did not show a strong correlation between plasma PAI-1 and serum CRP
in patients with colorectal cancer. Serum CRP was found to be a Dukes independent prognostic
variable in this patient cohort, and was found to identify a subgroup of curatively resected patients
at risk for short survival.
Key Words: Colorectal cancer—Plasminogen activator inhibitor-1—C-reactive protein—Sur-
There is now general consensus that 5-FU-based ad-
juvant chemotherapy improves long-term survival of pa-
tients with colorectal cancer (CRC) Dukes stage C.1,2
Similar results have not been reported for patients with
Dukes stage B disease. However, the significance of the
previous studies including Dukes stage B patients may
Received March 16, 2000; accepted June 16, 2000.
From the Departments of Surgical Gastroenterology (HJN, FM) and
Clinical Biochemistry (SS), Hvidovre University Hospital, Hvidovre;-
The Finsen Laboratory (IJC, NB), Rigshospitalet, University of
Copenhagen, Copenhagen; and The RANX05 Colorectal Cancer Study
Group (see Appendix).
Address correspondence to: Hans Jørgen Nielsen, MD, DMSc, Dept. of
Surgical Gastroenterology 435, Hvidovre University Hospital, DK-2650
Annals of Surgical Oncology, 7(8):617–623
Published by Lippincott Williams & Wilkins © 2000 The Society of Surgical Oncology, Inc.
have been limited by the small number of patients stud-
ied,3which could have masked a potential beneficial
effect. To select CRC patients for adjuvant therapy, the
search for prognostic markers to predict patients at risk
of recurrent disease has been intensified.4,5Such a selec-
tion of high-risk patients not only would spare a signif-
icant number of patients from the often toxic adjuvant
therapy, but also would reduce the number of patients
needed and the follow-up time required for reaching
valid conclusions from adjuvant treatment studies.
It is well established that proteolytic enzymes produced
by cancer cells or cells in the tumor stroma are involved in
tissue remodeling, which accompanies cancer cell invasion
and metastasis.6–8Among several enzyme systems ex-
pressed in cancer tissue, urokinase-catalyzed plasminogen
activation is thought to play a key role in tissue degrada-
tion,8,9activation of pro-metalloproteinases,10activation of
cytokines,11and angiogenesis,8,12all of which could lead to
increased metastatic potential of the cancer cells. Thus, the
content of urokinase plasminogen activator (uPA) in ex-
tracts from CRC tissue appears to be increased compared to
normal large bowel tissue from the same patients,13,14and
high content of uPA in tumor tissue has been shown to be
a marker of poor prognosis in CRC.15In addition, high
levels of the soluble uPA receptor (uPAR) recently were
shown to be a Dukes independent prognostic marker of
poor prognosis in CRC.5,9Because high levels of uPA and
uPAR were associated with poor prognosis, it was assumed
that high levels of the plasminogen activator inhibitor
type-1 (PAI-1) would be a marker of good prognosis. Sur-
prisingly, several studies have now shown that in various
cancer types high tumor tissue levels of PAI-1 are associ-
ated with short patient survival, and we have previously
published that high preoperative plasma PAI-1 levels are
associated with short survival of patients undergoing resec-
tion for CRC.16In accordance with these clinical data, we
recently showed in an animal model that absence of PAI-1
prevented cancer invasion and impaired tumor angiogene-
Presence of solid tumors may be associated with in-
creased plasma levels of the unspecific, inflammatory-
related acute phase protein C-reactive protein (CRP),18
and CRP levels have been shown to correlate with tumor
operability.19,20In addition, recent reports from small
patient populations have indicated that CRP is a Dukes
dependent prognostic marker in CRC.21,22
It has been suggested that PAI-1 might be a marker of
the acute phase response and thereby be associated with
CRP,23and even modulated by similar cytokine mecha-
nisms.18,24As described earlier in this paper, recent ex-
perimental evidence has suggested, however, that PAI-1
might have functions16that are clearly separated from
those described for CRP. The purpose of the present
study was, therefore, to compare preoperative plasma
PAI-1 and serum CRP levels in patients undergoing
elective resection for CRC, and to evaluate the prognos-
tic value of the substances.
PATIENTS AND METHODS
Patients scheduled for elective operation of CRC were,
placebo controlled, clinical study comparing the effect of
the histamine-2 receptor antagonist ranitidine versus pla-
cebo on long-term survival. All patients were without clin-
ical symptoms of infectious disease, and none had been
treated with systemic steroids, histamine-2 receptor antag-
onists, antibiotics, antiviral and/or cytotoxic agents within
two weeks prior to study entry. Patients with previous solid
tumors, patients with severe concurrent illness, such as
leukemia, autoimmune disease or HIV infection, and pa-
tients who were undergoing chemotherapy, radiotherapy, or
any other adjuvant therapy for cancer or were expected to
undergo such therapies within the study period were not
included. Patients unable to comply with the requirements
of the protocol were not allowed to participate, and patients
with a life expectancy shorter than 3 months at the time of
surgery were not included in the study. All eligible patients
received ranitidine or placebo from just before skin incision
and continued on medication for 5 years or until death.25
The numbers of patients receiving preoperative blood trans-
fusion were recorded.26The patients were followed every
Serum and EDTA plasma samples were collected
from all patients in the morning just before surgery. All
blood samples were collected before the first infusion of
ranitidine/placebo to avoid any potential confounding
effect of ranitidine. The samples were stored at ?80°C
until analysis. Plasma PAI-1 was determined by a sand-
wich ELISA method, including two monoclonal PAI-1
antibodies.27This assay detects both latent and active
PAI-1 and PAI-1:uPA complexes. Serum CRP was de-
termined using the nephelometric method, and concen-
tration of CRP was expressed as nmol/L; 25 nmol/L
being the detection limit.
The SAS software package (version 6.12; SAS Insti-
tute, Cary, North Carolina) was used to manage patient
data and perform statistical analyses. Clinical data such
as age, gender, Dukes stage of disease (using Dukes
original classification and addition of a D group identi-
fying patients with distant metastasis), tumor location,
specific perioperative blood component for transfused
patients, and overall survival were recorded. Rank sta-
tistics were used to calculate correlation coefficients and
618NIELSEN ET AL.
Ann Surg Oncol, Vol. 7, No. 8, 2000
to test hypotheses on location. Tests of independence
were done using the ?2test. The significance level was
set to 5%. The Kaplan-Meier method was used to esti-
mate survival probabilities, and the log-rank test was
used to test equality of strata. The end point was death,
of any cause. The association of plasma PAI-1 and CRP,
respectively, and survival was tested using the median
value of PAI-1 and the upper limit of the normal range
(94 nmol/L) of CRP. In addition, the PAI-1 and CRP
measurements were log transformed (natural log) for
survival analyses. For graphical presentation of survival,
the patients were stratified into groups based on CRP
values, so that each stratum yielded an equal number of
events (deaths). The Cox proportional hazards model
was applied for analysis of continuous covariates as well
as for multivariate analysis. The assumption of propor-
tional hazards was verified graphically.
Ranitidine had no effect on overall survival or survival
stratified according to Dukes classification,25and, there-
fore, treatment was not included in the statistical calcu-
lations. Samples from 594 patients were available for
plasma PAI-1 and serum CRP analyses. The median age
distribution was 70 years (range 33–89 years). The dis-
tribution of patients according to Dukes classification
was: A: 57; B: 223; C: 176; and D: 138. The median
follow-up period was 6.8 (5.4–7.9) years. In a univariate
survival analysis Dukes stage of disease appeared to be a
significant (P ?.0001, log-rank test) predictor of sur-
vival, indicating that the present patient population in
this respect was representative of CRC in general.
The median value of PAI-1 was 35.8 (range 5.2–269.0)
ng/ml; values ?94 nmol/L identified patients with in-
creased CRP levels. Comparison of PAI-1 and CRP levels
within individual patients showed a weak correlation of r ?
.26, P ? .0001. The median (range and 25%–75% quar-
tiles) distribution of plasma PAI-1 in the four Dukes groups
is shown as a box plot in Figure 1. There was no difference
in the median levels among patients with Dukes A, B, and
C disease, but patients with Dukes D disease had a signif-
icantly (P ? .045) higher level. The distribution of serum
CRP is shown in Figure 2. The median CRP levels in-
creased with Dukes stage of disease (P ? .0001), and the
highest values were found among patients with dissemi-
nated disease. Plasma PAI-1 levels were dependent on
tumor location, with a median (range) value of 37.4 (4.2–
347.6) ng/ml among patients with colon cancer (n ? 341),
and 33.7 (4.7–520.4) ng/ml among patients with rectal
cancer (n ? 253; P ? .02). The corresponding levels of
serum CRP were 129 (25–3,797) nmol/L and 49 (25–1902)
nmol/L, respectively (P ? .0001).
We have shown previously that by using the median
value of PAI-1 in a univariate survival analysis with a
median observation period of 2.1 (range 1.1–3.3) years,
high levels of PAI-1 were associated with poor prognosis
and low levels with good prognosis.16Performing uni-
variate survival analysis with the extended observation
period confirmed that high levels of PAI-1 were associ-
ated with poor prognosis and low levels with good prog-
nosis (P ? .02; HR: 1.3; 95% CI: 1.0–1.6; Fig. 3).
Considering PAI-1 as a continuous variable did not in-
dicate a statistically significant progressive effect on
survival (P ? .14). Similarly, high levels of CRP were
associated with poor prognosis and low levels with good
prognosis (P ? .0001; HR: 1.9; 95% CI: 1.6–2.4). Con-
sidering CRP as a continuous variable showed that in-
creasing levels were associated with poorer survival.
This is illustrated graphically by grouping patients by
their CRP levels as follows: those within the normal
range (? 94 nmol/L) formed one group, and those with
CRP levels of 95 nmol/L or higher were stratified into
four groups according to CRP level so that each stratum
plasma PAI-1 concentrations in 594 patients with primary colorectal
cancer. Each box shows the lower, median, and upper quartile; the
whiskers show the range. There was no difference in the median levels
among patients with Dukes A, B, and C disease, but patients with
Dukes D disease had a significantly (P ? .045) higher level.
Box plots of the individual measurements of preoperative
619PAI-1 AND CRP IN COLORECTAL CANCER
Ann Surg Oncol, Vol. 7, No. 8, 2000
had an equal number of events (and, thereby, similar
statistical power). We found a significant (P ?.0001)
association of increasing CRP level with continuously
increasing risk of mortality (Fig. 4).
Including only PAI-1 and CRP in a Cox proportional
hazards model showed that PAI-1 was not significant
(P ? .74, HR: 1.0), whereas CRP was highly significant
(P ? .0001; HR: 1.4; 95% CI: 1.3–1.5 [log scale]). In a
multivariate statistical analysis including Dukes classifi-
cation, age, gender, blood transfusion, and tumor loca-
tion, PAI-1 was found to be a Dukes-dependent prog-
nostic variable (logePAI-1; P ? .14, HR: 0.9; 95% CI:
0.7–1.0, Table 1). However, CRP was shown to be
Dukes-independent (logeCRP P ? .0001; HR: 1,4; 95%
CI: 1.3–1.5, Table 1). High levels of CRP identified
patients with poor prognosis; low levels identified pa-
tients with good prognosis. Moreover, patients with rec-
tal cancer had a shorter survival period than did patients
with colon cancer (P ? .0001; HR: 1.6; 95% CI: 1.3–2.1,
Table 1). Age scored in years at study entry also was
significantly associated with survival (P ? .0001),
whereas gender was not. Similar statistical analyses were
performed in the subgroup of patients who underwent
curative resection, differentiating between all curatively
resected patients (Dukes A ? B ? C) and exclusively
node-negative patients (Dukes A ? B). These separate
calculations still showed preoperative serum CRP as a
Dukes-independent prognostic variable (P ? .0001; HR:
1.3; 95% CI: 1.2–1.5) and (P ? .0009; HR: 1.3; 95% CI:
Three hundred seventy-two patients received periop-
erative blood component transfusion and in the univari-
ate statistical analysis transfusion predicted patients with
poor prognosis (P ? .04; HR: 1.3, 95% CI: 1.0–1.6)
compared with nontransfused patients. Perioperative
blood component transfusion had no independent prog-
nostic value in the multivariate analysis (P ? .52).
We have previously reported preoperative plasma
PAI-1 levels to be associated with survival of patients
with CRC.16By extending the median observation period
from 2.1 years to 6.8 years, the prognostic information of
plasma PAI-1 was confirmed. PAI-1 previously has been
suggested as a marker of the acute-phase response28and
correlated to the prototypical acute-phase protein CRP in
nonmalignant pathological conditions.28,29It has been
suggested, therefore, that levels of plasma PAI-1 might
be related to the acute-phase response in malignant dis-
eases. By analyzing serum CRP in the same blood sam-
serum CRP concentrations in 594 patients with primary colorectal
cancer. Each box shows the lower, median, and upper quartile; the
whiskers show the range. The horizontal solid line indicates the upper
limit of the normal range (94 nmol/L). The median CRP levels in-
creased with Dukes stage of disease (P ?.0001), and the highest values
were found among patients with disseminated disease.
Box plots of the individual measurements of preoperative
formed values) levels and overall survival of 594 patients with primary
colorectal cancer. By using the median value of 35.8 ng/ml the patients
were divided into two groups. The number of patients at risk is given
with 12-month intervals up to 48 months. In univariate analysis high
levels of PAI-1 were associated with poor prognosis and low levels
with good prognosis. The HR and P values are shown in the figure.
The association of preoperative plasma PAI-1 (logetrans-
620 NIELSEN ET AL.
Ann Surg Oncol, Vol. 7, No. 8, 2000
ples used for the PAI-1 determinations, it was found that
CRP levels increased with the disease stage; the highest
values were found among patients with advanced dis-
ease. This finding is similar to our findings with PAI-1.16
Comparison of the two molecules showed, however, that
there was only a weak correlation (r ? .26; P ?.0001).
This is in agreement with other studies in which a similar
lack of correlation between the two molecules was dem-
onstrated.30In addition, the results indicate, at least in
patients with CRC, that CRP determinations in serum are
unable to substitute PAI-1 determinations in plasma from
the same patients, as previously suggested.31
In a variety of solid tumors the tissue PAI-1 concen-
tration is higher than in the adjacent normal tissue.32
Moreover, in situ hybridization for PAI-1 showed a
strong expression of PAI-1 mRNA in endothelial cells in
tumor vessels in contrast to normal colon tissue, which,
like most other normal tissues, are devoid of detectable
PAI-1 mRNA.33These experiments suggest that at least
part of the PAI-1 found in CRC patients may be derived
from the tumor tissue. The platelet count is increased in
most cancer types, including CRC, and has been shown
to represent a prognostic variable.34Platelet-derived
PAI-1 also may contribute, therefore, to the high levels
of plasma PAI-1 found in patients with CRC. It has been
shown that PAI-1 accumulates in various blood compo-
nents for transfusion.35,36Patients with CRC included in
the present study who needed preoperative, intraopera-
tive, or postoperative blood transfusion received only
buffy-coat-depleted erythrocyte suspensions of SAGM
(saline-adenine-glucose-mannitol) blood.35This particu-
lar blood component contains ?1% of the amount of
platelets donated by the donor, and, therefore, the extra-
cellular level of PAI-1 is virtually undetectable.35
Experiments in wild-type mice and in mice in which
the PAI-1 gene has been disrupted show that lack of
expression of PAI-1 prevents tumor cell invasion and
impairs tumor angiogenesis.17In addition, PAI-1 has
been shown to compete with the urokinase receptor for
binding to the extracellular matrix protein vitronectin,
and it has been suggested that a surplus of PAI-1 will
release uPAR-expressing cells from the extracellular ma-
trix.9These effects of PAI-1 are clearly separated from
any known function of CRP.37
The increased levels of CRP in serum obtained preoper-
atively from patients with colorectal cancer may reflect the
inflammatory response stimulated by certain cytokines, in-
cluding interleukin-6 (IL-6), presumably released from var-
ious leukocytes due to the malignant process.28However,
infiltration of leukocytes, both the total cell count and
counts of specific cells, into the tumor tissue has been
associated with good prognosis and low levels with poor
prognosis.38IL-6 appears to play a dominating role in
levels may predict poor prognosis in various cancers, in-
cluding CRC.39,40Recent results have indicated that in-
creased IL-6 levels in cancer may be related to synthesis
and release by tumor cells,40and may reflect their prolifer-
classification, age, gender, tumor location, perioperative
blood transfusion, PAI-1, and CRP
Multivariate analysis including Dukes
P valueHR Lower level Upper level
Dukes B vs. A
Dukes C vs. A
Dukes D vs. A
CRP, C-reactive protein; HR, hazard ratio; PAI-1, plasminogen
values) levels and overall survival of 594 patients with primary colo-
rectal cancer. The patients were divided into one group with values
below the upper limit of the normal range (?94 nmol/L), and into four
strata with a comparable amount of events (deaths). The log-rank test
revealed a statistically significant difference (P ?.0001) of survival in
the four strata. CRP values 95–155 nmol/L ? stratum 1; 156–298
nmol/L ? stratum 2; 299–574 nmol/L ? stratum 3; ?574 nmol/L ?
stratum 4. The number of patients at risk is given with 12-month
intervals up to 48 months. The P value is shown in the figure.
The association of preoperative serum CRP (logetransformed
621 PAI-1 AND CRP IN COLORECTAL CANCER
Ann Surg Oncol, Vol. 7, No. 8, 2000
ative activity.41Thereby, increased synthesis and release of
IL-6 may lead to increased preoperative circulating CRP
levels, which may have prognostic value, as shown in the
present study of approximately 600 patients. It is still un-
known how IL-6 and, thereby, CRP may influence regula-
tion of tumor cell growth, but the molecules may have
multiple functions in health and in various disease process-
es.28It previously was suggested that perioperative treat-
ment with ranitidine may reduce surgery-induced IL-6-
associated CRP synthesis.42However, in the present study
therefore may not confound the results.
In our preliminary report we showed, in a univariate
associated with a poor prognosis in CRC.16The median
observation period was 2.1 years. These results were con-
firmed in the present study, where the observation time was
extended to 6.8 years. In the multivariate analysis, however,
plasma PAI-1 was found to be a Dukes-dependent and
serum CRP an independent prognostic factor. This under-
than in adjacent normal tissue.32Moreover, the indepen-
dence of serum CRP may further support the well-estab-
lished notion of the molecule as a reflection of the systemic
be directly associated with the extend of tumor burden. In
addition, the results may underline the weak correlation
between plasma PAI-1 and serum CRP in cancer, although
there may be a closer association between the molecules in
other pathological states, such as in patients with severe
The results of the present study underline the well-
accepted multifaceted biology of tumor cell growth, and
the difficulties of using only one prognostic marker to
detect patients with poor prognosis. This is further ac-
centuated by the finding that patients with rectal cancer
had an overall poorer prognosis compared to patients
with colon cancer, although the median levels of PAI-1
and CRP were significantly lower among patients with
rectal cancer. It is still unknown why levels of the mol-
ecules were significantly lower in patients with rectal
cancer. Significant differences in disease severity with
more node-negative diagnoses among patients with rec-
tal cancer versus patients with colon cancer may, in part,
explain the finding. Future studies of prognostic markers
should consider including both clinical parameters and
tumor markers with known biological activity into arti-
ficial neural network analyses44,45to identify patients not
only with Dukes C, but also with Dukes B disease at risk
for recurrent disease and poor survival. Thereby, the
overall outcome following treatment of patients with
CRC may possibly be improved, and, in addition, the use
of prognostic stratification would reduce the number of
patients required for reaching valid conclusions from
adjuvant treatment studies of early CRC.
The study received financial support from The Ingeborg
Roikjer Foundation, The Aage and Johanne Louis-Hansen
Foundation, The Danish Pharmacy Foundation of 1991, The
Agnete and Poul Friis Fund, The Walter and O. Kristiane
Christensen Fund, The E. Danielsen Family Fund, The Hede-
Nielsen Foundation, The Einar and Vera Hansen Foundation,
The Torben and Alice Frimodt Fund, and The Danish Cancer
Society (grant # 97 100 21).
The following persons participated in The Danish
RANX05 Colorectal Cancer Study Group:
Svend Schulze, MD, DMSc, Jens Thorup, MD, Peer
Wille-Jørgensen, MD, DMSc, Bispebjerg Hospital:
Copenhagen; Erik Bentzen, MD, Bornholm Hospital:
Bornholm (it’s an island hospital); Lars Banke, MD,
DMSc, Dorthe Froberg, MD, Frederikssund Hospital:
Frederikssund; Finn W. Henriksen, MD, DMSc, Poul
Crone, MD, Gentofte Hospital: Gentofte; Peter Hesself-
eldt, MD, Bo Hempel-Sparsø, MD, Karen Lindorff
Larsen, MD, Glostrup Hospital: Glostrup; Torsten As-
mussen, MD, Jørgen Heiner, MD, Helsingør Hospital:
Heisingør; Ole Hart Hansen, MD, DMSc, Henrik Flyger,
MD, PhD, Per Jess, MD, DMSc, Hillerød Hospital: Hill-
erød; Jørgen Iversen, MD, Jørgen LaCour Andersen,
MD, DMSc, Bo Vennits, MD, Holbæk Hospital: Hol-
bæk; Janne H. Hammer, MD, Hvidovre Hospital:
Hvidovre, Denmark; Anders Fischer, MD, DMSc, Hanne
Galatius, MD, Hørsholm Hospital: Hørsholm; Lars
Naver, MD, Kalundborg Hospital: Kalundborg; Dorthe
Teilum, MD, Køge Hospital: Køge; Leif Holbraad, MD,
Nykøbing Falster Hospital: Nykøbing Falster; Ole
Iversen, MD, DMSc, Jens Nymark, MD, Ole Roikjær,
MD, Næstved Hospital: Næstved; Leif Palm MD, DMSc,
Kirsten C. Rasmussen, MD, Roskilde Hospital: Rosk-
ilde; Jørn Friis, MD, Charlotte Lanng, MD, Henrik
Ovesen, MD, Slagelse Hospital: Slagelse; Niels Chr.
Jensen, MD, Niels Hoffman, MD, Stege Hospital: Stege;
Torben Larsen, MD, Jørgen Packler, MD, Sundby Hos-
1. Steele GD, Augenlicht LH, Begg CB, et al. National Institute of
Health consensus development conference statement: adjuvant ther-
apy for patients with colon and rectal cancer. JAMA 1990;264:352–8.
2. Casillas S, Pelley RJ, Milsom JW. Adjuvant therapy for colorectal
cancer. Dis Colon Rectum 1997;40:977–92.
622 NIELSEN ET AL.
Ann Surg Oncol, Vol. 7, No. 8, 2000
3. Moertel CG, Fleming TR, MacDonald JS, et al. Levamisole and
fluorouracil for adjuvant therapy of resected colon carcinoma.
N Engl J Med 1990;322:352–8.
4. Lindmark G, Gerdin B, Påhlman L, Bergstrøm R, Glimelius B.
Prognostic predictors in colorectal cancer. Dis Colon Rectum 1994;
5. Stephens RW, Nielsen HJ, Christensen IJ, et al. Plasma urokinase
receptor levels in patients with colorectal cancer: relationship to
prognosis. J Natl Cancer Inst 1999;91:869–74.
6. Danø K, Behrendt N, Bru ¨nner N, Ellis V, Ploug M, Pyke C. The
urokinase receptor. Protein structure and role in plasminogen ac-
tivation, and cancer invasion. Fibrinolysis 1994;8:189–203.
7. Hewitt R, Danø K. Stromal cell expression of matrix-degrading pro-
tease systems in human cancer. Enzyme Protein 1996;43:163–73.
8. Mazar AP, Henkin J, Goldfarb RH. The urokinase plasminogen
activator system in cancer: Implications for tumor angiogenesis
and metastasis. Angiogenesis 1999;3:15–32.
9. Bru ¨nner N, Nielsen HJ, Hamers M, et al. The urokinase plasmin-
ogen activator receptor in blood from healthy individuals and
patients with cancer. APMIS 1999;107:160–7.
10. Murphy G, Atkinson S, Ward R, Gavrilovic J, Reynolds JJ. The
role of plasminogen activators in the regulation of connective
tissue metalloproteinases. Ann N Y Acad Sci 1992;667:1–12.
11. Odekon LE, Blasi F, Rifkin DB. Requirement for receptor-bound
urokinase in plasmin-dependent conversion of latent TGF-beta to
active TGF-beta. J Cell Biol 1994;158:398–407.
12. Mignatti P, Rifkin DB. Plasminogen activators and matrix metal-
loproteinases in angiogenesis. Enzyme Protein 1996;49:117–37.
13. Sier CFM, Vloedgraven HJM, Ganesh S, et al. Inactive urokinase
and increased levels of its inhibitor type 1 in colorectal cancer liver
metastasis. Gastroenterology 1994;107:1449–56.
14. Sier CFM, Vloedgraven HJM, Griffioen G, et al. Plasminogen acti-
vators and inhibitor type 1 in neoplastic colonic tissue from patients
with familial adenomatous polyposis. Br J Cancer 1995;71:393–6.
15. Verspaget HW, Sier CFM, Ganesh S, Griffioen G, Lamers C
BHW. Prognostic value of the plasminogen activators and their
inhibitors in colorectal cancer. Eur J Cancer 1995;31A:1105–9.
16. Nielsen HJ, Pappot H, Christensen IJ, et al. Association between
plasma PAI-1 levels and survival in patients with colorectal cancer.
17. Bajou K, Noe ¨l A, Gerard R, et al. Absence of host plasminogen
activator inhibitor-1 prevents cancer invasion. Nature Med 1998;
18. Ballou SP, Kushner I. C-reactive protein and the acute phase
response. Adv Intern Med 1992;37:313–36.
19. Durdey P, Williams NS, Brown DA. Serum carcinoembryonic
antigen and acute phase proteins in the preoperative detection of
fixation of colorectal tumours. Br J Surg 1984;71:881–4.
20. Go ¨ransson J, Jonsson S, Larsson P. Pre-operative plasma levels of
C-reactive protein, albumin and various plasma protease inhibitors
for the pre-operative assessment of operability and recurrence in
cancer surgery. Eur J Surg Oncol 1996;22:607–17.
21. McMillan DC, Wotherspoon HA, Fearon KCH, Sturgeon CM,
Cooke TG, McArdle CS. A prospective study of tumour recurrence
and the acute phase response after apparently curative colorectal
cancer. Am J Surg 1995;170:319–22.
22. Nozoe T, Matsumata T, Kimamura M, Sugimachi K. Significance
of elevation of serum C-reactive protein as an indicator for prog-
nosis in colorectal cancer. Am J Surg 1998;176:335–8.
23. Kluft C, Verheijen JH, Jie AFH, et al. The postoperative fibrino-
lytic shutdown: a rapidly reverting acute phase pattern for fast-
acting inhibitor of tissue-type plasminogen activator after trauma.
Scand J Clin Invest 1985;45:605–10.
24. Whawell SA, Thompson JN. Cytokine-induced release of plasmin-
ogen activator inhibitor-1 by human mesothelial cells. Eur J Surg
25. Nielsen HJ, McArdle CS, Moesgaaard F, et al. The effect of
ranitidine on long-term survival in primary colorectal cancer. A 40
months interim analysis. GI Cancer 1998;2:227–33.
26. Nielsen HJ. Clinical impact of bioactive substances in blood trans-
fusion components. Implications for leukofiltration. Infuther
Transfusion Med 1998;25:296–304.
27. Grøndahl-Hansen J, Christensen IJ, Rosenquist C, et al. High
levels of urokinase-type plasminogen activator and its inhibitor
PAI-1 in cytosolic extracts of breast carcinomas are associated
with poor prognosis. Cancer Res 1993;53:2513–21.
28. Gabay C, Kushner I. Acute-phase proteins and other systemic
responses to inflammation. N Engl J Med 1999;340:448–54.
29. Haverkate F, Thompson SG, Duckert F, et al. Haemostasis factors
in angina pectoris; relation to gender, age and acute-phase reaction,
results of the ECAT Angina Pectoris Study Group. Thromb He-
30. Kirchheimer JC, Ko ¨lbl H, Tatra G, Binder BR. Relationship be-
tween plasma levels of components of the fibrinolytic system and
acute-phase reactants in patients with uterine malignancies. Eur
J Clin Invest 1990;20:79–84.
31. Sattar N, McMillan DC. Association between plasma plasminogen
activator inhibitor-1 and survival in colorectal cancer. BMJ 1998;317:
32. Pappot H, Gårdsvoll H, Rømer J, et al. Plasminogen activator
inhibitor type 1 in cancer: therapeutic and prognostic implications.
Bio Chem Hoppe-Seyler 1995;376:259–67.
33. Pyke C, Kristensen P, Ralfkier E, Eriksen J, Danø K. The plas-
minogen activation system in human colon cancer: messenger
RNA for the inhibitor PAI-1 is located in the endothelial cells in
tumor stroma. Cancer Res 1991;51:4067–71.
34. Monreal M, Fernandez-Llamazares J, Pinol M, et al. Platelet count
and survival in patients with colorectal cancer-a preliminary study.
Thromb Haemost 1998;79:916–8.
35. Nielsen HJ, Reimert CM, Pedersen AN, et al. Time-dependent,
spontaneous release of white cell- and platelet-derived bioactive
substances from stored human blood. Transfusion 1996;36:960–5.
36. Edvardsen L, Taaning E, Mynster T, Hvolris J, Drachman O,
Nielsen HJ. Bioactive substances in buffy-coat-derived platelet
pools stored in platelet-additive solutions. Br J Haematol 1998;
37. Deng G, Curriden SA, Wang SJ, Rosenberg S, Loskutoff DJ. Is
plasminogen activator inhibitor-1 the molecular switch that gov-
erns urokinase receptor-mediated cell adhesion and release? J Cell
38. Nielsen HJ, Hansen U, Christensen IJ, Reimert CM, Bru ¨nner N,
Moesgaard F. Independent prognostic value of eosinophil granu-
locytes and mast cells in colorectal cancer tissue. J Pathol 1999;
39. Goydos JS, Brumfield AM, Frezza E, Booth A, Lotze MT, Carty
SE. Marked elevation of serum interleukin-6 in patients with
cholangiocarcinoma: validation of utility as a clinical marker. Ann
40. Piancatelli D, Romano P, Sebastiani P, Adorno D, Casciani CU.
Local expression of cytokines in human colorectal carcinoma:
evidence of specific interleukin-6 gene expression. J Immunother
41. Kinoshita T, Ito H, Miki C. Serum interleukin-6 level reflects the
tumor proliferative activity in patients with colorectal carcinoma.
42. Rasmussen LA, Nielsen HJ, Sørensen S, et al. Ranitidine reduces
postoperative interleukin-6 induced C-reactive protein synthesis.
J Am Coll Surgeons 1995;181:138–44.
43. Inthorn D, Hoffmann JN, Hartl WH, Muhlbayer D, Jochum M.
Effect of antithrombin III supplementation on inflammatory re-
sponse in patients with sever sepsis. Chock 1998;10:90–96.
44. Bottaci L, Drew PJ, Hartley JE, et al. Artificial neural networks
applied to outcome prediction for colorectal cancer patients in
separate institutions. Lancet 1997;350:469–72.
45. Burke HB, Goodman PH, Rosen DB, et al. Artificial neural net-
works improve the accuracy of cancer survival prediction. Cancer
623PAI-1 AND CRP IN COLORECTAL CANCER
Ann Surg Oncol, Vol. 7, No. 8, 2000