Background and Aim: The introduction of antihormonal
and anti-inflammatory targeted therapy have been reported with various success
rates in recurrent and locally advanced desmoid tumors. The aim of the current
study is to evaluate the immunohistochemical detection of COX-2 and ER? in desmoid
tumors and assess their correlation with available clinicopathologic variables.
Material and Methods: A
total of 17 desmoid tumor cases (11 abdominal, 5 extra-abdominal and 1
intra-abdominal) were examined for immunohistochemical detection of COX-2 and
ER? using monoclonal antibodies. Toluidine blue staining was performed to
confirm or exclude that COX-2 immunostained cells coincide with mast cells in
co-localized sections. Correlation of results with available clinicopathologic
variables was done and p value < 0.05 was considered significant. Results: COX-2 was expressed in tumor cells in 92% of examined desmoid cases (16/17). Toluidine blue staining revealed that COX-2 immunostained cells don't coincide with the few metachromatically stained mast cells in co-localized sections. ER? was expressed in 67.1% of tumor cells in desmoid cases (11/17); 8 cases displayed high ER? expression and 3 cases displayed low ER? expression. No significant correlation was detected between ER? or COX-2 immunohistochemical expression and patient's age, gender, tumor size, site, margins status and recurrence history (p>0.05).
Conclusions: This study confirmed the
immunohistochemical expression of COX-2 and ER? in tumor cells of the majority
of studied desmoid cases. These results introduce COX-2 and ER? as potential
therapeutic targets in desmoid tumors. Further studies with large sample size
and follow up are recommended to validate the current results.
Key words: COX-2, ER?, desmoid, target
tumors are locally infiltrative soft tissue growth with high rate of local
recurrence even after total surgical resection 1.The current main treatment strategy
of desmoid tumors or as known aggressive fibromatosis is surgical excision
accompanied by radiotherapy, but the failure to achieve complete response after
surgery and the possible morbidity linked to radiotherapy in some cases has
emerged the need for investigating the role of possible target pharmacological therapy
high incidence of desmoid tumors in females (about 80%) and the high rate of
reported cases during or following pregnancy and spontaneous regression in
postmenopause have raised the possibility of hormonal factors as major players
in the pathogenesis of desmoid tumors 4.
detection of estrogen receptors in desmoid tumors was reported in early studies
using ligandbinding assays; however, more current studies using
immunohistochemical techniques yielded conflicting results regarding expression
of ER? and ER? 5.
wider distribution of ER? than ER? in several tissues as prostate, thyroid, and
mesenchymal tissues and the recent availability of specific antibodies to ER?
directed researchers to investigate its role in pathogenesis of desmoid tumors
as most studies reported absence of immunohistochemical expression of ER? 5.
enzyme family is involved in prostaglandins production from arachidonic acid.
COX-2 is one of the inducible family members that is overexpressed in different
types of cancer. It enhances tumorigenesis through inhibiting apoptosis,
induction of angiogenesis, invasive potential, and stimulating growth factors
such as platelet-derived growth factors (PDGF). Accordingly, nonsteroidal
anti-inflammatory drugs (NSAIDs) that inhibits COX-2 activity were introduced
as target therapy that have beneficial effects in cancer prevention 2, 6.
NSAIDs as indomethacin, have been examined either single or in association with
other hormonal treatment such as tamoxifen in treatment of desmoid tumors 4, with reported partial or total response rates ranging from 37%
to 57% in multiple studies 7.
aimed in the current study to evaluate the immunohistochemical detection of COX-2 and
ER? in desmoid tumors and assess their correlation with available clinicopathologic
Material and Methods
this study, a total of 17 archived paraffin blocks of desmoid tumors were
retrieved retrospectively from Pathology department, faculty of medicine, Cairo
University in the period from January 2016 till June 2017. All samples included
were obtained through total surgical excision. The margins were inadequate in 8
cases. Twelve patients were females, while 5 were males. The cases were
classified according to their site as abdominal, intra- abdominal and extra-abdominal.
Of note, the intra- abdominal desmoid tumor included in this study was mesenteric
mass associated with familial adenomatous polyposis (FAP). Fifteen cases
included were primary desmoids and two cases were recurrent.
17 retrieved formalin fixed paraffin embedded tissue blocks were cut at 5-?m-thickness,
stained by H&E and examined microscopically to confirm the diagnosis of
Immunohistochemical staining of COX-2 and ER?:
sections were cut at 5-?m-thickness from each of the 17
paraffin embedded sections of desmoid cases on immunostaining slides. Applying
the standard protocol of Dako; heat mediated retrieval of antigen was performed
by applying citrate buffer pH 6 in automated water bath (Dako PT link, PT101).
The primary antibodies were COX2 monoclonal antibody (SP21), #MA5-1456,
manufactured by Thermofisher Scientific, USA and monoclonal ER beta
(PPG5/10),#MA1 81281, manufactured by Thermofisher Scientific, USA. An
autostainer (Dakoautostainer link 48) was used for immunostaining using a
polymer-based detection system (DakoEnVision™ FLEX, K8000). Diaminobenzidine
(DAB) was applied as chromogen and counterstaining was done using Mayer’s haematoxylin.
Afterwards, the cover slips and DPX were used for mounting and preserving
tissue sections. The positive control used for COX-2 & ER beta was
urothelial carcinoma positive for COX-2 granulosa cells in normal ovary
section was cut from the paraffin blocks at 5 -?m-thickness,
stained with toluidine blue to highlight mast cells.
Olympus BX51 light microscope equipped with a digital camera was used for
examination and capturing of digital images.
Evaluation of immunostaining of both COX-2 & ER?:
of COX-2 was reported as positive if more than 10% of tumor cells showed cytoplasmic
positivity for COX-2 8.The
ER? nuclear immunostaining was reported as negative (< 5% positive tumor cells ), low expression (5-25% positive tumor cells) , high expression (>
25% positive tumor cells) 9.
results of COX-2 and ER? immunostaining were correlated with multiple
clinicopathologic factors (age, sex, tumor size, margin status and recurrence
Evaluation of toluidine blue stained sections:
COX-2 positive immunostained cells were compared with metachromatically stained mast cells in
co-localized touldine blue stained sections to reveal if they coincide or not.
analysis of all results and available variables were performed using The
Statistical Package for the Social Sciences (SPSS) version 15. The Chi-square
test was used to assess difference between qualitative variables. P values less
than 0.05 levels were considered to be statistically significant.
study took the approval of ethical committee in faculty of medicine, Cairo
total of 17 cases of desmoid tumors were studied (figure 1). The patient’s age
displayed wide range from 7 years up to 50 years old with median age 33 years.
Most patients were females (70.6%) while 29.4% were males. Tumors more than 6
cm in greatest dimension constituted 64.7% of the cases (6/17). The
localization of the tumors was classified as abdominal (11/17), extra-abdominal
(5/17) and intra-abdominal (1/17). All desmoid tumors enrolled in this study
were treated by surgical excision. All clinicopathologic data are summarized in
nuclear staining was expressed in 64.7% (11/17) of cases; 47.1% (8/17)
displayed high expression (figure 2) and 19.6% (3/17) displayed low expression
(figure 3), while 35.3% (6 /17) of cases showed negative staining.
The majority of the studied cases
(16/17) showed positive cytoplasmic immunohistochemical staining for COX-2,
while only one case showed negative staining.
COX-2 positive immunostained cells were similar in morphology to tumor cells in
haematoxylin and eosin stained sections. The toluidine blue stained co-localized
sections highlighted only few mast cells with specific metachromatic staining
that didn’t coincide with COX-2 positive immunostained cells as shown in figure
4A and 4B.
significant correlation was detected between ER? and COX-2 expression with
different clinicopathologic variables including patient’s age, gender, tumor
size, site, margins status and history of recurrence as displayed in table 2
and 3 respectively.
immunohistochemical expression was not significantly correlated with ER?
expression as shown in table 4.
tumors are rare tumors with incidence rate 2 to 4 per million annually 10. Although
, the traditional treatment of desmoid tumors is surgery, the infiltrative
borders of these tumors and the absence of tumor capsule makes complete
surgical resection with negative margins not always successful 11, 12.
introduction of systemic therapy as antihormonal and anti-inflammatory agents have been reported
with various success rates in recurrent, unresectable and locally advanced
desmoid tumors 13.
studies reported that COX-2 inhibitors induced the shrinkage of desmoid tumors
in clinical trials as COX-2 induces the growth of desmoid tumors through
increased prostaglandin E2 production14, 15, 16. Moreover,
complete regression of desmoid tumors with tamoxifen was reported in other
studies17 ,18, achieving
recurrence free period up to 9 years in study performed by Maseelall et al 19.
aimed in this work to examine the expression of COX-2 and ER? and in desmoid
tumours using immunohistochemistry as it is practical accessible and cost effective method available
in almost all surgical pathology laboratories. To best of our knowledge,
several studies assessed ER? detection in desmoid tumors by
immunohistochemistry but only two studies examined COX2 expression and
contradictory results were reported 14, 16.
the current study, tumor cells expressed COX-2 protein in 92% (16/17 cases) of
desmoid tumors without statistical significant correlation with all
clinicopathologic variables including age, gender, tumor site, size, recurrence
and margin status (p< 0.05). These results coincide with Signoroni et al., results as all of their 14 studied desmoid tumors showed COX-2 positive immunohistochemical expression in tumor cells and their results were confirmed by presence of COX-2 mRNA by means of RT-PCR 14. In contrast, Emori et al. reported that 56% of their 16 studied desmoid cases expressed COX-2 in mast cells only as clearly co-expressed by tryptase. This was against our findings which were confirmed by toluidine blue staining and proved that COX-2 immnnostained cells don't coincide with few mast cells in co localized sections 16.This controversy in results of COX-2 expression may be owed to different antibodies used and small sample size enrolled in different studies. In addition, COX-2 expression was examined by Poon and his colleagues in cell cultures of 33 cases of aggressive fibromatosis and proved to be positive in 31 cases through RT-PCR and in only 3 out of 17 cases of normal tissue dissected from resection margins. Furthermore, the levels of PGE2 were higher in cell cultures of aggressive fibromatosis than examined normal margins fibroblasts (p>0.05) 20.
was expressed in 64.7% of our studied desmoid tumor cases (11/17); high
expression was reported in 8 cases (47.1%) and low expression was reported in 3
cases (19.6%). Six cases showed negative ER? expression (35.3%). The
correlation between ER? expression and all clinicopathologic variables
including age, gender, tumor site, size, recurrence, margin status and COX 2
expression as well was non-significant (p> 0.05). Similar figures were
reported by Zhang et al., 9.as
80.5% of their studied desmoid tumors were positive for immunohistochemical expression
of ER?; most (67.5%) showed high-level expression, and 13% showed low-level expression.
investigators reported positive ER? expression in almost majority of their cases
as Deyrup et al. 5.showed
that 100% of their studied extraabdominal desmoid tumors were positive for ER?.
Santos et al., 21,also
stated that 89% of their desmoid tumor cases showed positivity for ER? without
statistical significant correlation with all clinical parameters.
contrast, Leithner et al. 22, reported
low expression of ER? in their desmoid tumors (7/80); four out of 46 cases in
extra-abdominal, two out of 21 in abdominal, and one out of 13 in
intraabdominal cases. Ishizuka et al., 23, also
reported low percentage of ER? expression in two out of 27 desmoid tumors
(7.4%). Different antibodies used and different sample size may contribute to
this diversity in reported results.
sum up, we reported high percentage of COX-2 and ER? immunohistochemical
expression in tumor cells of the examined desmoid tumors. These observations,
recommend antihormonal therapy and COX 2 inhibitors as possible therapeutic
targets in management of desmoid tumors. More prospective trials with larger
sample size and follow up after treatment are needed to validate these results.