C-reactive protein inhibits

lymphangiogenesis and resultant

lymph node metastasis of squamous
cell carcinoma in mice
Tomohiko Sasaki, MD, PhD,a Satoru Motoyama, MD, PhD,a Yusuke Sato, MD, PhD,a
Kei Yoshino, MD, PhD,a Goichi Matsumoto, PhD,b Yoshihiro Minamiya, MD, PhD,a
Hajime Saito, MD, PhD,a Katsuyuki Murata, MD, PhD,c and Jun-ichi Ogawa, MD, PhD,a
Akita and Yokosuka, Japan

Background. Lymph node involvement is the most important prognostic factor in many solid cancers.
Recently, we found that patients with esophageal and lung cancer carrying the C-reactive protein (CRP)
1846T/T genotype, which is associated with lower serum CRP levels, are more likely to have lymph node
metastasis. We hypothesized that host CRP directly inhibits lymph node metastasis.
Methods. We inoculated NR-S1M metastatic cells subcutaneously into the backs of C3H/HeN mice.
Concurrently, 1 mg of recombinant mouse CRP or phosphate-buffered saline was injected subcutaneously
every 3 days for 5 weeks, after which the mice were killed for evaluation. We evaluated lymph node
metastasis and lymphangiogenesis in the implanted tumor by using immunohistochemical analysis with
anti-pancytokeratin and antilymphatic vessel endothelial hyaluronan receptor-1 antibodies.
Results. There was no substantial difference in tumor size between the 2 groups but the lymph nodes were
smaller in the CRP group than the control group (P < .044). Immunohistochemical analysis confirmed
inguinal lymph node metastasis in 70% (14/20) of control mice, but in only 30% (3/10) of mice in the
CRP group. Moreover, the metastatic area within lymph nodes was less in the CRP group (P < .042)
and tumoral lymphangiogenesis was decreased in the CRP group (P < .037).
Conclusion. CRP appears to inhibit tumoral lymphangiogenesis and lymph node metastasis in mice.
These findings suggest that by inhibiting lymph node metastasis, CPR may have therapeutic potential for
use against cancer. (Surgery 2013;154:1087-92.)

From the Departments of Surgerya and Environmental Health Sciences,c Akita University Graduate School of
Medicine, Akita; and Department of Oral Surgery,b Kanagawa Dental College Clinical Education Center,
Yokosuka, Japan

LYMPH NODE INVOLVEMENT is the most important
prognostic factor in many cancers, which makes
it essential that nodal metastases are targeted
when treating solid cancers.1 Furthermore, lymph
node metastasis does not always correlate with fac-
tors present in the primary neoplasm; indeed, mul-
tiple and distant lymph node metastases are often
Supported, in part, by Grants-in-Aid for Scientific Research
from the Ministry of Education, Culture, Science, Sports and
the Japan Science and Technology Agency.
Accepted for publication April 11, 2013.
Reprint requests: Satoru Motoyama, MD, PhD, Department
of Surgery, Akita University Graduate School of Medicine,
1-1-1 Hondo, Akita 010-8543, Japan. E-mail: motoyama@doc.
med.akita-u.ac.jp.
0039-6060/$ - see front matter
2013 Mosby, Inc. All rights reserved.
http://dx.doi.org/10.1016/j.surg.2013.04.016
associated with relatively small tumors. Thus, a bet-
ter understanding of the molecular and biologic
mechanisms governing lymph node metastasis
would be highly desirable.
Several reports have shown that increased se-
rum levels of C-reactive protein (CRP) are associ-
ated with cancer progression and a poor
prognosis.2-6 In addition, we found that lymph
node involvement in esophageal or lung cancer
may have a genetic component, ie, the CRP
1846C>T genetic polymorphism appears to be an
independent factor associated with lymph node
metastasis.7,8 For example, patients with esopha-
geal cancer carrying the 1846T/T genotype, which
is associated with lesser serum CRP levels, are
greater than 3 times more likely to have lymph
node involvement.
Before metastasis, primary tumors generally
exhibit lymphangiogenesis, with new lymphatics

SURGERY 1087
1088 Sasaki et al
extending from the tumor to the tumor-draining
lymph nodes. This observation suggests that the
key factors mediating lymph node metastasis are
present at the tumor site. We hypothesized that
CRP produced by the host may serve to inhibit
tumoral lymphangiogenesis, thereby inhibiting
lymph node metastasis. To test that idea, we
examined the effect of CRP on tumoral lymphan-
giogenesis and consequent lymph node metastasis
and whether CRP has the potential to serve as an
anticancer agent inhibiting lymph node
metastasis.
MATERIALS AND METHODS
Tumor cells. The NR-S1M murine squamous
cell carcinoma cell line was used.
These cells arose spontaneously in the C3H/He
mouse strain and demonstrate lymph nodes me-
tastases.9 The cells were cultured in RPMI-1640
(Sigma-Aldrich, St. Louis, MO) supplemented
with 10% heat-inactivated fetal bovine serum
(GIBCO, Grand Island, NY) and antibiotics (100
U/mL penicillin G, 100 mg/mL streptomycin,
250 mg/mL amphotericin B; GIBCO) in a humidi-
fied incubator at 378C under 5% CO2/95% air.
Animal model of lymph node metastasis. Fe-
male C3H/HeN mice (6 8 weeks old) were
obtained from CLEA (Tokyo, Japan). All animal
experiments were performed in accordance with
the policies of the animal ethics committee of
Akita University Graduate School of Medicine.
Using a syringe with a 27-gauge needle, we
subcutaneously inoculated NR-S1M metastatic cells
(5 3 106 cells) into the backs, near the tail, of
C3H/HeN mice under ether anesthesia. At the
same time, some mice (CRP group; n = 10) were
administered 1 mg of recombinant mouse CRP (r-
CRP; R&D Systems, Minneapolis, MN) in 100 mL
of PBS subcutaneously into their backs near their
neck. In the control group, 100 mL of PBS was in-
jected subcutaneously into the backs near the neck
(n = 20). Thereafter, the 2 groups were given PBS
or r-CRP every 3 days for 5 weeks (12 injections of
CRP or PBS). The mice were then killed, and the
tumors and inguinal lymph nodes were extracted.
We examined the inguinal lymph nodes for metas-
tasis because they drain the part on the back where
the tumor cells were injected.9 Tumor and lymph
node volumes were calculated with the following
formula: volume (mm3) = length (mm) 3 width
(mm)2 3 1/2. For analysis, each tumor was divided
into 2 tissue samples; one was stored frozen at
808C, whereas the other was fixed in 10% para-
formaldehyde. All lymph nodes were fixed in
10% paraformaldehyde.
Surgery
November 2013
Serum CRP concentration in mice. Blood sam-
ples were collected from the jugular vein on day 0
(pretreatment with r-CRP), 6, 12, or 18, whereas
the mice were under anesthesia. Those mice were
then killed (n = 3 in each day). Serum CRP con-
centrations were determined using a Mouse C-re-
active Protein Detection Kit ELISA (Helica
Biosystems, Fullerton, CA) according to the manu-
facturers protocol.
Evaluation of lymph node metastasis. After cut-
ting the fixed lymph nodes into 4-mm-thick sections,
the sections were deparaffinized in xylene and eth-
anol, placed in 10 mmol/L Tris buffer (pH 9)
containing 1 mmol/L ethylenediaminetetraacetic
acid, and irradiated via a microwave (750 W) for 5
minutes. Endogenous peroxidase activity was
blocked by incubating the sections for 15 minutes
in 3% H2O2, and nonspecific binding was blocked by
incubation for 30 minutes in 10% goat serum
(Nichirei, Tokyo, Japan). The lymph node sections
were then incubated for 60 minutes with mouse mon-
oclonal anti-pancytokeratin (AE1/AE3) antibody
(1:200 dilution, sc-81714; Santa Cruz Biotechnology,
Santa Cruz, CA) as the primary antibody. This was fol-
lowed by incubation in blocking buffer and then with
a peroxidase-conjugated, anti-mouse antibody
(MAX-PO(Mouse); Histofine Mousestain Kit; Ni-
chirei) for 10 minutes each. The color development
reagent used was diaminobenzidine (Nichirei), and
the slides were reacted for 5 minutes. All reactions
were conducted at room temperature. We defined
AE1/AE3-positive lymph nodes as metastatic.
We examined the AE1/AE3-positive areas in each
lymph node section at 403 magnification. Digital
images of each lymph node stained with AE1/AE3
were examined, and the AE1/AE3-positive area,
expressed as percentage of the total lymph node
area, was calculated using ImageJ software (Na-
tional Institutes of Health, Bethesda, MD).
Evaluation of lymphangiogenesis in the im-
planted tumor. Tumor sections (4 mm) were incu-
bated first for 60 minutes with rabbit, polyclonal
anti-lymphatic vessel endothelial hyaluronan
receptor-1 (LYVE-1) antibody (1:200 dilution,
ab14917; Abcam, Cambridge, UK) and then for
10 minutes with a peroxidase-conjugated, anti-
rabbit antibody (MAX-PO(Rabbit); Histofine
Simple Stain Mouse). Details of the immunohisto-
chemistry are the same as described previously.
Lymphangiogenesis was evaluated by counting the
LYVE-1 positive lymphatic vessels within the tu-
mor tissue at 2003 magnification under a light
microscope. In addition, digital images of all
tumors stained for LYVE-1 were captured, and
the LYVE-1 positive areas per unit area of tumor
Surgery
Volume 154, Number 5
section (mm2/1 3 108 mm2) were calculated using
ImageJ software.
The frozen tumor samples were lysed in Qpro-
teome Mammalian Protein Prep Kit (QIAGEN,
Dusseldorf, Germany) using a TissueLyser II (QIA-
GEN) according to the manufacturers protocol.
The protein contents of the samples were then
measured by the use of standard bicinchoninic
acid assays, after which aliquots of lysate contain-
ing 10 mg of protein were subjected to 12.5%
sodium dodecyl sulfate polyacrylamide gel electro-
phoresis and transferred to a polyvinylidene di-
fluoride membrane (Atto, Tokyo, Japan).
After blocking the membrane for 60 minutes with
3% skim milk in Tris-buffered saline containing
0.1% Tween 20, it was incubated for 60 minutes
with rabbit polyclonal anti-LYVE-1 antibody (1:500
dilution, sc-28190; Santa Cruz Biotechnology). The
probed membrane was then incubated for 60 min-
utes with goat, anti-rabbit IgG-HRP (1:5000 dilution,
sc-2004; Santa Cruz Biotechnology) and exposed to
enhanced chemiluminescence reagents (sc-2048;
Santa Cruz Biotechnology) followed by radiographic
exposure and development. After each step, the
membrane was washed 3 times for 5 minutes in Tris-
buffered saline containing 0.1% Tween 20, and all
reactions were conducted at room temperature. We
performed this experiment 3 times, each time cap-
turing digital images and calculating the intensity of
the LIVE-1 signal using ImageJ software (National
Institutes of Health, Bethesda, MD). The relative
intensity of the LIVE-1 band in the CRP group was
expressed as a ratio with the control group (CRP/
control) in each experiment.
Assessing distant metastasis. Five weeks after
inoculation of the NR-S1M cells, distant organ
metastases were assessed with an in vivo, micro X-
ray computed tomography system (R_mCT2, Ri-
gaku, Tokyo, Japan) in 3 anesthetized mice.
Statistical analysis. Values are expressed as the
median [interquartile range]. Differences between
the 2 groups were analyzed using the Wilcoxon
signed-rank test. Differences in the metastasis rate
between the control and CRP groups were assessed
using Fishers exact test. Statistical analysis was
performed using JMP 9 (SAS Institute, Cary, NC).
RESULTS
The serum CRP level increased from 3.8
[3.6 4.5] ng/mL (day 0) to 5.7 [5.4 6.1] ng/
mL on day 6 after initiating subcutaneous injec-
tions of recombinant CRP. Thereafter, the level
continued to increase gradually, reaching 9.3
[7.0 10.5] ng/mL on day 12 and 12.4
[8.6 12.5] ng/mL on day 18 (Fig 1).
Sasaki et al 1089
Fig 1. Model of lymph node metastasis in mice. After im-
plantation of NR-S1M cells, mice were subcutaneously
injected with recombinant-CRP (1 mg) every 3 days.
Bars depict median values along with the interquartile
ranges of the serum CRP concentrations (ng/mL) in
mice administered recombinant CRP (n = 3 in each
day). (A) Tumor formation (arrow) became detectable
on day 7 after implantation of cancer cells. (B) The tu-
mor (arrow) and swollen left inguinal lymph node (circle)
5 weeks after implantation of the cancer cells.
Tumor formation was first observed on day 7
after implantation of NR-S1M cells in all control
and CRP mice. Five weeks after tumor cell implan-
tation, there was no difference in tumor size
between the 2 groups (tumor volume was 3.2
[1.9 4.1] 3 103 mm3 in the control group and
2.3 [1.6 3.0] 3 103 mm3 in the CRP group, P =
.301; Fig 2, A). Affected lymph nodes were not de-
tectable from the outward appearance of the mice.
After making an incision on the back, however, we
found subcutaneous inguinal lymph nodes. The
total number of resected lymph nodes was 27 in
the control group and 15 in the CRP group, and
the median number of lymph nodes resected per
mouse was 1 [1 2] in the control group and 1.5
[1 2] in the CRP group (P = .452). The volume
of the lymph nodes was 1.6 [0.6 2.7] mm3 in the
CRP group and 3.7 [1.0 9.0] mm3 in the control
group. Thus, lymph nodes were smaller in the
CRP group than the control group (P < .044;
Fig. 2, B).
CT confirmed that none of the mice showed
other organ metastasis. Metastases to the inguinal
lymph nodes was seen in 70% (14/20) of control
mice, whereas the incidence of lymph node metas-
tasis in the CRP group was only 30% (3/10; P =
.056). In addition, the median percent metastatic
area was 13 [0 63]% in the control group, but
only 0 [0 9]% in CRP group mice (P < .042; Fig 3).
In Figure 4, A and B, tumors were stained for
LYVE-1. The LYVE-1 positive area per unit area of
tumor section was 5.4 [3.07.1] mm2/1 3 108 mm2
1090 Sasaki et al Surgery
November 2013

Fig 2. (A) Tumor volumes 5 weeks after implantation of tumor cells. Control: n = 20, CRP: n = 10. (B) Lymph node
volumes 5 weeks after implantation of tumor cells. Control: n = 20 mice, 27 lymph nodes, CRP: n = 10 mice, 15 lymph
nodes. Bars depict median values along with the interquartile ranges.

Fig 3. Detection of lymph node metastasis in inguinal lymph nodes through immunohistochemical staining for pancy-
tokeratin (AE1/AE3). Metastatic cancer cells are stained brown. (A) and (B) Lymph node without metastasis. (C) and
(D) Lymph node with metastasis. Scale bars are 500 mm in (A) and (C) and 50 mm in (B) and (D). (D) Also shows cancer
cells within a lymphatic vessel near the lymph node. Arrows show a lymph vessel that contains cancer cells.

in the control group and 2.3 [0.04.6] mm2/1 3 108
mm2 in the CRP group (P <.037, Fig 4, C). Consistent
with that finding, Western blot analysis showed the
overall level of LYVE-1 protein present within tumors
from CRP mice to be less than in tumors from control
mice (Fig. 4, D). The relative intensities of the LIVE-
1 bands in the CRP group (CRP/Control) were 0.20,
0.27, and, 0.48, respectively.
DISCUSSION
In this study we showed for the first time that CRP
inhibits tumoral lymphangiogenesis and the
apparently resultant lymph node metastasis in
mice. To date, there have been 3 published reports
on the effects of CRP on tumor metastasis.10-12 They
suggest that CRP may function to control distant or-
gan metastasis in mice bearing malignant tumors. In
contrast, there have been no reports investigating
whether CRP can stimulate lymph node metastasis
in a mouse model. Control of lymph node metastasis
is a target for therapy for some solid cancers, and in-
hibiting lymph node metastasis may require control-
ling tumoral lymphangiogenesis. In the present
study, we used immunohistochemistry with an anti-
Surgery Sasaki et al 1091
Volume 154, Number 5

Fig 4. Immunohistochemical staining of intratumor lymphatic vessels for LYVE-1. Arrows show lymphatic vessels that
stained light brown. (A) Tumor from a control mouse. (B) Tumor from a CRP mouse. Scale bars are 50 mm. (C) Quan-
tification of lymphangiogenesis. LYVE-1 positive areas per unit area of tumor section were evaluated (control: n = 20,
CRP: n = 10). Bars depict median values along both with the interquartile ranges. (D) Western blots showing tumoral
expression of LYVE-1 in a control and a CRP mouse.

LYVE-1 antibody to evaluate intratumor lymphan- disease by impairing cardioprotective prostanoids.26

giogenesis. LYVE-1 is a reliable marker of lymphatic
vessels that is expressed predominantly in lymphatic
endothelial cells,13,14 making it useful in the analysis
of lymphangiogenesis associated with lymph node
metastasis.15,16 In our study, LYVE-1 positive areas
within tumor sections were significantly less in
CRP-treated mice than control mice, suggesting
that host CRP may decrease intratumor lymphangio-
genesis and thus suppress lymph node metastasis.
Several growth factors affecting lymphangiogenic
activity have been reported. In a study of cutaneous
squamous cell carcinomas in vascular endothelial
growth factor (VEGF)-A transgenic mice, VEGF-A
induced active proliferation of tumor-associated
lymphatic endothelial cells and could potentially
serve as a tumoral lymphangiogenesis factor.17 The
effect of VEGF-A was to attract macrophages to tu-
mors, which are sites of inflammation.18,19 The
tumor-associated macrophages then promoted in-
tratumor lymphangiogenesis by abundantly produc-
ing VEGF-A, VEGF-C, and VEGF-D.19-23
CRP was identified originally as a biomarker of
acute and chronic inflammation.24,25 In the context
of cardiovascular disease, CRP reportedly reduces ni-
tric oxide production, causes tyrosine nitration of
endothelial PGI2 synthase, and inhibits PGI2 release
from arterioles. Thus, CRP promotes endothelial
dysfunction and the development of coronary artery
Cyclooxygenase-2 (COX-2) is also expressed in re-
gions of inflammation. COX-2 is the predominant
source of PGI2, so that COX-2 inhibitors, such as eto-
dolac, impair the production of PGI2 and predis-
pose one to cardiovascular disease.27,28 In contrast,
COX-2 inhibitors also suppress lymphangiogenesis
in vivo.18 Cox-2 suppresses immune responses and
promotes cancer cell invasion and angiogenesis by
stimulating the secretion of proangiogenic factors,
including VEGF-A and fibroblast growth factor-2,
from cancer cells and/or stromal fibroblasts.29-32
Hirakawa33 suggested the novel concept of the
lymphovascular niche to explain expansion of lym-
phatic networks. This concept suggests that primary
tumors are themselves able to actively modify future
metastatic sites for preferential relocation and to po-
tentially enlarge the sinusoidal lymphatic endothe-
lium for the migration of cancer cells to the lymph
nodes. We suggest that CRP may function within
the lymphovascular niche to prevent lymph node me-
tastasis. Moreover, because the occurrence of lymph
node metastasis involves the concerted actions of a
number of factors, lymph node metastasis may only
be suppressed when several factors are targeted.
In esophageal and renal cancers, tumors ex-
pressing CRP are associated with a poor progno-
sis.34 We suggest that the function of tumoral CRP
differs from that of host CRP. Whereas tumoral
1092 Sasaki et al
CRP acts to extend cancer cells, host CRP inhibits
metastasis, including lymph node metastasis. Fur-
ther studies will be required to explore these ideas.
REFERENCES
1. Gourin CG, Johnson JT. Surgical treatment of squamous cell
carcinoma of the base of tongue. Head Neck 2001;23:653-60.
2. Pierce BL, Ballard-Barbash R, Bernstein L, Baumgartner
RN, Neuhouser ML, Wener MH, et al. Elevated biomarkers
of inflammation are associated with reduced survival among
breast cancer patients. J Clin Oncol 2009;21:3437-44.
3. Katriina H, Shah E, Debbie A. A systematic review of the as-
sociation between circulating concentrations of C-reactive
protein and cancer. J Epidemiol Community Health 2007;
61:824-32.
4. Beer TM, Lalani AS, Lee S, Mori M, Eilers KM, Curd JG,
et al. C-reactive protein as a prognostic marker for men
with androgen-independent prostate cancer: results from
the ASCENT trial. Cancer 2008;112:2377-83.
5. Kristine H, Stig E, Brge G. Baseline C-reactive protein is as-
sociated with incident cancer and survival in patients with
cancer. J Clin Oncol 2009;27:2217-24.
6. Koike Y, Miki C, Okugawa Y, Yokoe T, Toiyama Y, Tanaka K,
et al. Preoperative C-reactive protein as a prognostic and
therapeutic marker for colorectal cancer. J Surg Oncol
2008;98:540-4.
7. Motoyama S, Miura M, Hinai Y, Maruyama K, Usami S, Saito
H, et al. CRP genetic polymorohism is associated with
lymph node metastasis in thoracic esophageal squamous
cancer. Ann Surg Oncol 2009;16:2479-85.
8. Minamiya Y, Miura M, Hinai Y, Saito H, Ito M, Imai K, et al.
The CRP1846T/T genotype is associated with a poor prog-
nosis in patients with non-small cell lung cancer. Tumor
Biol 2010;31:673-9.
9. Matsumoto G, Yajima N, Saito H, Nakagami H, Omi Y, Lee
U, et al. Cold shock domain protein A (CSDA) overexpres-
sion inhibits tumor growth and lymph node metastasis in a
mouse model of squamous cell carcinoma. Clin Exp Metas-
tasis 2010;27:539-47.
10. Deodhar D, James K, Chiang T, Edinger M, Barna BP. Inhi-
bition of lung metastasis in mice bearing a malignant fibro-
sarcoma by treatment with liposomes containing human
C-reactive Protein. Cancer Res 1982;42:5084-8.
11. Kresl JJ, Potempa LA, Anderson B, Radosevich JA. Inhibi-
tion of mouse mammary adenocarcinoma (EMT6) growth
and metastasis in mice by a modified form of C-reactive pro-
tein. Tumor Biol 1999;20:72-87.
12. Franco MA, Mendoza E, Zapata P, Castillo P, Isaza CE,
Zamora D, et al. Antiangiogenic and antitumor effects
of IMMUNEPOTYENT CRP in murine melanoma. Im-
munopharmacol Immunotoxicol 2010;32:637-46.
13. Tammela T, Saaristo A, Holopainen T, Lyytikka J, Kotronen
A, Pitkonen M, et al. Therapeutic differentiation and matu-
ration of lymphatic vessels after lymph node dissection and
transplantation. Nat Med 2007;13:1458-66.
14. Oliver G, Detmar M. The rediscovery of the lymphatic sys-
tem. Old and new insights into the development and bio-
logical function of the lymphatic vascular system. Genes
Dev 2002;16:773-83.
15. Banerji S, Ni J, Wang SX, Clasper S, Su J, Tammi R, et al.
LYVE-1, a new homologue of the CD44 glycoprotein, is
lymph-specific receptor for hyaluronan. J Cell Biol 1999;
144:789-801.
Surgery
November 2013
16. Jackson DG, Prevo R, Clasper S, Banerji S. LYVE-1, the lym-
phatic system and tumor lymphangiogenesis. Trends Immu-
nol 2001;22:317-21.
17. Hirakawa S, Kodama S, Kunstfeld R, Kajiya K, Brown LF,
Detmer M. VEGF-A induces tumor and sentinel lymph
node lymphangiogenesis and promotes lymphatic metasta-
sis. J Exp Med 2005;201:1089-99.
18. Iwata C, Kano MR, Komuro A, Oka M, Kiyono K, Johansson
E, et al. Inhibition of cyclooxgenase-2 suppresses lymph
node metastasis via reduction of lymphangiogenesis. Can-
cer Res 2007;67:10181-9.
19. Cursiefen C, Chen L, Borges LP, Jackson D, Cao J, Radzie-
jewski C, et al. VEGF-A stimulates lymphangiogenesis and
hemangiogenesis in inflammatory neovascularization via
macrophage recruitment. J Clin Invest 2004;113:1040-50.
20. Jeon BH, Jang C, Han J, Kataru RP, Piao L, Jung K, et al.
Profound but dysfunctional lymphangiogenesis via vascular
endothelial growth factor ligands from CD11b+ macro-
phages in advanced ovarian cancer. Cancer Res 2008;68:
1100-9.
21. Dadras SS, Paul T, Bertoncini J, Brown LF, Muzikansky A,
Jackson DG, et al. Tumor lymphangiogenesis: a novel prog-
nostic indicator for cutaneous melanoma metastasis and
survival. Am J Pathol 2003;162:1951-60.
22. Schoppmann SF, Birner P, Stockl J, Kalt R, Ullrich R, Caucig
C, et al. Tumor-associated macrophages express lymphatic
endothelial growth factors and are related to peritumoral
lymphangiogenesis. Am J Pathol 2002;161:947-56.
23. Baluk P, Tammela T, Ator E, Lyubynska N, Achen MG, Hick-
lin DJ, et al. Pathogenesis of persistent lymphatic vessel hy-
perplasia in chronic airway inflammation. J Clin Invest
2005;115:247-57.
24. Steven B, Irving K, David S. C-reactive protein. J Biol Chem
2004;279:48487-90.
25. Lorraine M, Carolyn M, Du Clos TW. C-reactive protein: lig-
ands, receptors and role in inflammation. Clin Immunol
2005;117:104-11.
26. Travis W, Erion Q, Yi R. C-reactive protein impairs coronary
arteriolar dilation to prostacyclin synthase activation: role of
peroxynitrite. L Mol Cell Cardiol 2009;47:196-202.
27. Solomon SD, Wittes J, Finn PV, Fowler R, Viner J, Bertag-
nolli MM, et al. Cardiovascular risk of celecoxib in 6 ran-
domized placebo-controlled trials: the cross trial safety
analysis. Circulation 2008;117:2104-13.
28. Grad E, Golomb M, Koroukhov N, Lawson JA, Lotan C, Fitz-
gerald GA, et al. Aspirin reduces the prothrombotic activity
of C-reactive protein. J Thromb Haemost 2009;7:1393-400.
29. Tsuji M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois RN.
Cyclooxygenase regulates angiogenesis induces by colon
cancer cells. Cell 1998;93:705-16.
30. Williams CS, Tsuji M, Reese J, Dey SK, DuBois RN. Host
cyclooxygenase-2 modulates carcinoma growth. J Clin Invest
2000;105:1589-94.
31. Dannenberg AJ, Altorki NK, Boyle JO, Dang C, Howe LR,
Weksler BB, et al. Cyclooxygenase 2: a pharmacological target
for the prevention of cancer. Lancet Oncol 2001;2:544-51.
32. Wang D, Mann JR, Dubois RN. The role of prostaglandins
and other eicosanoids in the gastrointestinal tract. Gastro-
enterology 2005;128:1445-61.
33. Hirakawa S. From tumor lymphangiogenesis to lymphovas-
cular niche. Cancer Sci 2009;100:983-9.
34. Nakatsu T, Motoyama S, Maruyama K, Usami S, Sato Y,
Miura M, et al. Tumoral CRP expression in thoracic esoph-
ageal squamous cell cancers is associated with poor out-
comes. Surg Today 2012;42:652-8.