Detection of early lymphangiogenesis by lymphatic microvascular density and endothelial...

Rapid Communication

Detection of early lymphangiogenesis by lymphaticmicrovascular density and endothelial proliferation status inpreneoplastic and neoplastic lesions of the uterine cervixpin_2673 395..400

Anca Maria Cimpean,1 Vitalie Mazuru,2 Anatolie Cernii,3 Raluca Ceausu,1 Lilian Saptefrati,2

Alexandru Cebanu,2 Ana Maria Fit1 and Marius Raica1

1Department of Histology, Victor Babes University of Medicine and Pharmacy, Timisoara, Romania and 2Departmentof Histology, Nicolae Testemitanu University of Medicine and Pharmacy, Chisinau, Republic of Moldova, 3Departmentof Pathology, Institute of Oncology, Chisinau, Republic of Moldova

The characteristics of lymphangiogenesis in preneoplasticand neoplastic lesions of the uterine cervix are not wellknown and the role of this process in tumor progressionand metastasis is not well understood. The aim of thepresent study was to characterize the morphology and dis-tribution of lymphatic vessels and lymphatic proliferativestatus and to evaluate the value of lymphatic microvasculardensity (LMVD) in premalignant and malignant lesions ofthe uterine cervix. One hundred and twenty-eight paraffin-embedded cervical specimens were immunostained withD2-40 antibody specific for lymphatic endothelial cells.Colocalization of D2-40 and Ki67 for the proliferative char-acterization of lymphatic vessels was obtained by perform-ing double immunostaining. A low density of lymphaticvessels was detected in normal cervix and squamous meta-plasia. Intense and particular lymphangiogenic responsewas found in low and high grade squamous intraepitheliallesions and microinvasive carcinoma. Lymphatic prolifera-tion occurred early in cervical lesions, being more active inpremalignant lesions and microinvasive carcinomas than ininvasive lesions. Our results suggest an early initiation ofan active lymphangiogenesis in cervical lesions. These find-ings support the hypothesis that cervical preneoplasticlesions represent a critical point in the development of thelymphatic network vasculature. Early lymphangiogenesiscould explain lymph node metastasis associated with cer-vical invasive carcinomas at preliminary diagnosis.

Key words: D2-40, lymphangiogenesis, lymphatic microvascu-lar density (LMVD), uterine cervix

Cervical neoplasia remains one of the most controversialissues for clinicians, pathologists and researchers. Screeningprograms have reduced the incidence of invasive neoplasticlesions but have not changed the rate of precursor lesions.HPV is confirmed to be involved in the etiology of uterinecervix lesions. Recently, Coutlée et al. reported that samplesfrom women with low grade squamous intraepithelial lesions(LSIL) contained higher HPV-6 loads than women withoutthis lesion.1 Despite the existence of an anti-HPV vaccine,HPV gene expression is dramatically altered during cervicalcarcinogenesis and novel biomarkers are needed for a bettercharacterization of neoplastic cervical progression.2 Thepresence of HPV is likely, but not necessary, for the devel-opment of precursors and neoplastic lesions of the uterinecervix. Few morphological and molecular data are availableabout other factors involved in these lesions.3

Lymphangiogenesis—the formation of new lymphaticvessels (LV)—is evaluated in various tumor types, such ascolon malignancies,4 breast cancer,5 and malignant tumors ofthe esophagus.6

Uterine cervical lesions and the involvement of LV in thetumor development, progression and metastasis still remainsa field of debate. Few articles describe lymphatic ves-sel density and the potential role of lymphangiogenesis inthe progression of cervical neoplasia.7,8 The relationshipbetween lymphangiogenesis and the invasion of cervicalcancer was suggested by Hashimoto et al., who consideredthat vascular endothelial growth factor (VEGF)-C could be apromoter of pelvic lymph node metastasis in invasive cervicalcancer.9

Usually, malignant lesions of the uterine cervix are consid-ered more important than precursor lesions. This may explainthe lack of data about the promotion of lymphangiogenesis inpathological conditions of the uterine cervix. No data aboutthe origin of lymph vessels, the prognostic impact of lym-phangiogenesis in precursor lesions of the uterine cervix, or

Correspondence: Anca Maria Cimpean, MD, PhD, Departmentof Histology, Victor Babes University of Medicine and Pharmacy,Piata Eftimie Murgu nr.2, 300041 Timisoara, Romania. Email:[email protected]

Received 17 June 2010. Accepted for publication 21 March 2011.© 2011 The AuthorsPathology International © 2011 Japanese Society of Pathology andBlackwell Publishing Asia Pty Ltd

Pathology International 2011; 61: 395–400 doi:10.1111/j.1440-1827.2011.02673.x

their involvement in nodal pelvic metastasis developmenthave been confirmed. Available data about LV in tumors ofthe uterine cervix are rare and controversial, and moreover,there are almost no references regarding lymphangiogenesisin precursor lesions and a normal cervix.10,11

These reasons described were enough to make us inves-tigate a broad spectrum of neoplastic lesions of the uterinecervix and to focus on the identification of LV. The pur-pose was to evaluate morphology, distribution, lymphaticmicrovessel density and lymphatic proliferation in differentstages of cervical lesions. Comparison of type-specific fea-tures of lymphangiogenesis in premalignant and malignantlesions, correlated with proliferative status of lymphaticendothelial cells, could help detect the early lymphangiogen-esis in pathological conditions of the uterine cervix. Acomplete characterization of the morphological and immuno-histochemical data of LV could have a prognostic and thera-peutic impact and confirm diagnosis in cervical lesions.

MATERIALS AND METHODS

We included in our study 128 retrospective targeted biopsiesof the uterine cervix and specimens taken from conization inpatients with macroscopically detectable lesions. Specimenswere fixed in buffer formalin and embedded in paraffin, basedon the conventional histological technique. Step sections,5 mm thick, were prepared for each case. Initial sections werestained with HE, for the pathological diagnosis and grade ofthe tumor. Lesions were stratified as follows: squamous cellmetaplasia (n = 27); cervical intraepithelial lesions (LSIL, n =25) and (high grade intraepithelial squamous lesions (HSIL),n = 23); carcinoma in situ (n = 10); microinvasive carcinoma(n = 15); and invasive carcinoma (n = 20). Normal uterinecervix specimens (biopsies collected from patients withinconclusive results of colposcopy when conization was rec-ommended by gynecologist) were used as control (n = 8).Additional sections from each case were stained for D2-40,which recognizes the formalin-insensitive epitope of podopla-nin. The immunohistochemical technique used was based onavidin biotin technique using LSAB+ working system whichfollowed incubation with primary antibody D2-40, RTU,(DakoCytomation, Glostrup, Denmark) for 30 min.

Colocalization of the Ki67 proliferation marker (monoclonalantibody, clone MIB1 from Dako, (Carpinteria, CA, USA))with D2-40 in lymphatic endothelium was obtained by apply-ing the double staining method followed by the use of twodifferent chromogens (3,3′diaminobenzidine for nuclearbrown staining of Ki67 and amino ethyl carbazole for cyto-plasmic red staining of D2-40 epitope). Basal cells of strati-fied squamous epithelium of the cervix and lymphaticendothelial cells were considered positive control. Nucleistaining was performed with Lillie’s modified haematoxylin.

The entire immunohistochemical procedure was performedwith a DakoCytomation Autostainer.

Lymphatic microvessel density (LMVD) was assessedbased on the hot spot method, using the following protocol:three hot spots from each section were chosen at low powermagnification and counting was performed at ¥200 magnifi-cation. The arithmetic media of the three fields was the finalresult. The counting followed all the steps recommended byWeidner et al.12 and Van der Auwera et al.13 Microscopicimages were captured as JPEG format, and the area of theLV was calculated using Nikon Lucia G program of analysisof the microscopic image (Nikon, Tokyo, Japan).

Assessment of D2-40/Ki67-positive endothelial LV wasperformed on the same microscopic areas chosen for lym-phatic microvessel density, using the same microscopicevaluation. From the total number of positive LV found ina chosen area, we then counted only the LV whichco-expressed D2 40/Ki67. Only the vascular structures withwell defined lumen lined by D2-40-positive endothelial cellswere counted and after that those which met this same cri-teria as well as with Ki67-positivity in the endothelium. Noother structures with D2-40- or D2-40/Ki67-positive cord likestructures or isolated lymphatic endothelial cells withoutevident lumen were assessed.

Statistical analysis was done using SPSS version 13.0(SPSS, Chicago, IL, USA). Analysis included the use of theSPSS Data Editor tests (means test (Student’s t-test)), partialcorrelation tests, and non-parametric tests (c2 test); P < 0.05was considered significant.

For graphic representation of the statistical analysis weselected a stacked bar chart model applied for summaries ofseparate variables.

The local research ethics committee approved the protocolof the study, and informed consent was obtained from allsubjects according to the World Medical Association Decla-ration of Helsinki.

RESULTS

Basal cells of the epithelium from normal exocervix and cer-vical squamous metaplasia were positive for D2-40 as acontinuous layer over the lamina propria of the cervix. Thisdistribution was also observed in LSIL. A lower intensity ofimmunostaining was detected for basal cells of HSIL. A dis-continuous pattern of the basal layer or clusters of positivecells also characterized HSIL. D2-40 was negative in endot-helial cells of blood vessels and strongly stained the lym-phatic endothelial cells. The reaction had membraneenhancement in the case of lymphatic endothelial cells and adiffuse cytoplasmic pattern was noticed in basal cells.

There were differences concerning the distribution anddensity of LV between the histopathological types of cervical

396 A. M. Cimpean et al.

© 2011 The AuthorsPathology International © 2011 Japanese Society of Pathology and Blackwell Publishing Asia Pty Ltd

Figure 2 Proliferative status of lymphatic vessels in various cervi-cal lesion types (D2-40 in red and Ki 67 brown immunostain). (a)High grade intraepithelial squamous lesions and (b) in situ carci-noma with a high number of Ki 67 proliferative endothelial cells(nuclear, brown) lining D2-40-positive lymphatic vessels (red). (c)Microinvasive carcinoma also presented proliferative lymphaticendothelial cells. (d) Intratumoral lymphatic vessel from invasivecarcinoma with proliferating lymphatic endothelial cells at thetip of the lymphatic sprout (arrow) and collapsed lumen lined byD2-40-positive lymphatic endothelial cells.

Figure 3 Statistical analysis of the quantitative assessment of the total number of lymphatic vessels (LyVsT) and proliferative lymphaticvessels (Ly VsP) quantified on double D2-40/Ki67 immunostained specimens for different lesions of the cervix from (a) squamous metapla-sia (SM), (b) low grade intraepithelial squamous lesions (LSIL) and (c) high grade intraepithelial squamous lesions (HSIL), (d) to microinvasive(MI) and (e) invasive (INV) carcinoma. Note that significant (Sig.) correlation of proliferative index appear from early stages of cervical lesions(pLSIL = 0.009) and persist in invasive carcinoma (pINV = 0.002).

Figure 1 Double immunostaining for lymphatic vessels (D2-40,red) and proliferation marker (Ki67, brown). Scattered, large lym-phatic vessels disposed deeply in the lamina propria of the normaluterine cervix (yellow arrows), far from the normal epithelium. Notethe expression of Ki67 proliferation marker in the basal and supra-basal layer of the cervical epithelium (a). D2-40 positive lymphaticvessel (red) with numerous sprouting processes close to LSIL (b).Plexiform-like structures with numerous intravascular pillars suggest-ing lymphatic intussusception (c). Many lymphatic vessels close to insitu carcinoma lesion (d).

Lymphangiogenesis in cervical lesions 397


lesions included in our study. In the normal cervix we foundlarge LV, round or oval in shape without branches, founddeep in the lamina propria, far from the stratified squamousepithelium. In the normal cervix, D2-40-positive lymphaticvessels had no positive reaction for Ki67 in endothelial cells(Fig. 1a). The same pattern was found in squamous meta-plasia of the uterine cervix. LMVD ranged between 4.8 and6.6 vessels/¥200 magnification, with an average of 5.8 in thenormal cervix. In squamous cell metaplasia the number of LVwas not significantly different from results found in the normalcervix. No significant correlation was found concerning pro-liferative status of lymphatic endothelial cells in squamousmetaplasia (Fig. 3a).

LSIL and HSIL were characterized by the presence of LVclose to the epithelial component. Two types of lymphaticvessels were observed in LSIL and HSIL. Tube-like struc-tures with narrow lumen and a high branching tendencywere predominant close to LSIL (Fig. 1b). Associatedplexiform-like organized LV in LSIL were noticed in theselesions as multiple associated luminal structures whichseemed to arise from the same lymphatic vessel (Fig. 1c).Within the lumen of these LV we observed D2-40-positivepillar structures that split the lumen in either a complete orincomplete fashion. A significant increase in the number ofLV was found in cases with HSIL and in situ carcinoma(Fig. 1d). LMVD values ranged between 3 and 21/¥200,with an average of 11. The same value was found for in situcarcinoma. We noted a lack of positive reaction for basalcells in carcinoma in situ. All cases with LSIL, HSIL andcarcinoma in situ showed D2/40+/Ki67+ proliferative lym-

phatic endothelial cells. Proliferative vessels were distrib-uted close to the lesions (Fig. 2a,b).

A significant correlation was found between the prolifera-tive status of lymphatic endothelial cells in both the cervicalintraepithelial neoplasia types. In LSIL a higher correlationwas found compared with HSIL (P = 0.009 in LSILcompared with P = 0.044 for HSIL, see Fig. 3b,c). Thisfinding supports early onset of lymphangiogenesis in cervi-cal lesions.

In some cases, D2-40-positive LV surrounded areas ofmicroinvasive carcinoma. LMVD ranged between 5 and 39vessels/¥200 with an average of 15.7. High proliferativeactivity of lymphatic endothelial cells found in LSIL, HSIL andin situ carcinoma was also present in microinvasive carci-noma (Fig. 2c). A significant correlation of lymphatic prolif-erative activity and the microinvasive type of cervical cancerwas also found (P = 0.002, Fig. 3d).

In invasive carcinoma, LV were observed in the tumor andperitumoral areas. A few intratumoral LV were found as smallD2-40-positive vessels with narrow or collapsed lumen,irregular walls and no tumor cells inside. Peritumoral LV weresignificantly more numerous, larger, sinuous, with a discon-tinuous wall and often containing proliferative lymphaticendothelial cells and tumor cells (Fig. 2d). Density of lym-phatics varied between 1 and 15.2 with an average of 7.25.A significant correlation was also evident for invasive carci-noma (P = 0.002, Fig. 3e).

A schematic representation of lymphatic endothelial prolif-eration in different cervical lesions using a scatter plot chart isshown in Fig. 4.

Figure 4 Distribution of the total numberof lymphatic vessels and proliferating lym-phatic vessels from squamous metaplasia(MS), to invasive carcinoma stages of cer-vical lesions. High lymphatic proliferationwas found starting from cervical intrae-pithelial lesions (low grade intraepi-thelial squamous lesions (LSIL) and highgrade intraepithelial squamous lesions(HSIL)) to invasive carcinoma. Lymphaticmicrovessels density assessed by D2-40(dotted line) gradually increased fromsquamous metaplasia to intraepitheliallesions (LSIL and HSIL) first and after thatfrom microinvasive to invasive carcinoma.Number of lymphatic proliferative vessels(solid line) had a similar distribution beinghigher for intraepithelial lesions and with aslightly decrease for invasive carcinoma.



DISCUSSION

Lymph node metastases represent an important prognosticfactor in women with cervical cancer. Despite lymph nodestatus not being included in the staging system for cervicalcancer, the preoperative assessment of lymph nodes byimaging or intraoperative assessment significantly influencesthe therapeutic decision and patient survival.14 At the time ofclinical diagnosis, early stage cervical cancers are oftenassociated with pelvic and para-aortic lymph node metasta-sis.15 In microinvasive carcinoma of the uterine cervix,Hasumi et al. reported that the number of lymph node withmetastasis increases with the depth of invasion anddescribed ‘capillary-like’ involvement around microinvasivearea also related with the depth of invasion.16 The authorsnoted that no patients with such capillary-like spaces aroundmicroinvasive areas had nodal metastasis. These findingsare supported by our results which demonstrated that thesestructures were positive for D2-40 and have a high LMVD inmicroinvasive carcinoma of the uterine cervix.

Using the method of step sections, Roche and Norris,17

and later Leman et al.,18 demonstrated lymphatic invasion inmicroinvasive carcinoma without association of lymph nodemetastasis. Later, Zivaljeviç et al. reported similar resultsand recommended an individualized treatment based on anexhaustive pathological evaluation of an adequate conebiopsy specimen.19 These findings support the hypothesisthat in microinvasive carcinoma the newly formed LV(based on a higher LMVD found in the present study com-pared with normal cervix) are already functional. Theincrease of LMVD in the early stages of LSIL and thesimilar LMVD value found in the present study in HSIL,LSIL, and carcinoma in situ could partially explain the earlydevelopment of a lymphatic network and may promote earlynodal metastases in cervical cancer.

Scattered data characterizes lymphangiogenesis in cer-vical intraepithelial neoplasia lesions. Lymphangiogenicgrowth factors VEGF-C and VEGF-D, and their correspond-ing receptor VEGFR3 mRNA levels, significantly increase inHSIL lesions compared with LSIL.8 Correlation of these datawith our findings concerning a doubled LMVD value in LSILand HSIL compared to the normal cervix suggests an earlylymphangiogenic switch in cervical carcinogenesis.

Present findings concerning lymphangiogenesis correlatedwith our previous published data about proliferation of endot-helial cells from blood vessels in cervical lesions20 show anearly lymphatic endothelial proliferation in preneoplasticstages of cervical lesions even before the development of theangiogenic response. Based on these findings we hypoth-esize that the tumor lymphatic network is already organizedand functional when the cervical lesions become invasive.This could explain the presence of nodal metastasis in manycases at the time of invasive cancer diagnosis.

Intussusceptive microvascular growth is a fast process thatcan occur within hours or even minutes because it does notneed proliferation of endothelial cells. This process is widelyaccepted as a distinct angiogenic mechanism necessary forrapid expansion and remodeling of the blood capillary bed,21

but is less described for lymphatic capillaries.22 Our resultssuggest that lymphangiogenesis in preneoplastic and neo-plastic lesions of the uterine cervix is governed by bothsprouting and intussusceptive mechanisms also involved inangiogenic process.

In summary, the present study suggests by microscopy,immunohistochemistry and statistical analysis that lymphan-giogenesis is an early event in the lesions of the uterinecervix before an overt invasion certified by morphological andimmunohistochemical tests. This explains the early develop-ment of metastasis from microinvasive carcinoma stage,aggressiveness and poor response to the therapy of cervicalinvasive carcinoma. Further studies are necessary to evalu-ate angiogenesis and lymphangiogenesis in preneoplasticlesions of the uterine cervix to show their value for prognosisand therapeutic management.

ACKNOWLEDGMENT

This work was supported by grants PN II 41-054 and IDEI1147/2009 of Romanian Ministry of Education and Research.

REFERENCES

1 Coutlée F, Trottier H, Gagnon S et al. Low-risk human papillo-mavirus type 6 DNA load and integration in cervical samplesfrom women with squamous intraepithelial lesions. J Clin Virol2009; 45: 96–9.

2 Brandsma JL, Sun Y, Lizardi PM et al. Distinct human papillo-mavirus type 16 methylomes in cervical cells at different stagesof premalignancy. Virology 2009; 389: 100–7.

3 Vinokurova S, Wentzensen N, Kraus I et al. Type-dependentintegration frequency of human papillomavirus genomes in cer-vical lesions. Cancer Res 2008; 68: 307–13.

4 Duff SE, Jeziorska M, Kumar S et al. Lymphatic vessel density,microvessel density and lymphangiogenic growth factor expres-sion in colorectal cancer. Colorectal Dis 2007; 9: 793–800.

5 Liu HT, Ma R, Yang QF et al. Lymphangiogenetic characteristicsof triple negativity in node-negative breast cancer. Int J SurgPathol 2009; 17: 426–31.

6 Saad RS, Lindner JL, Liu Y et al. Lymphatic vessel density asprognostic marker in esophageal adenocarcinoma. Am J ClinPathol 2009; 131: 92–8.

7 Longatto-Filho A, Pinheiro C, Pereira SM et al. Lymphaticvessel density and epithelial D2-40 immunoreactivity in pre-invasive and invasive lesions of the uterine cervix. GynecolOncol 2007; 107: 45–51.

8 Van Trappen PO, Steele D, Lowe DG et al. Expression of vas-cular endothelial growth factor (VEGF)-C and VEGF-D, andtheir receptor VEGFR-3, during different stages of cervical car-cinogenesis. J Pathol 2003; 201: 544–54.

Lymphangiogenesis in cervical lesions 399


9 Hashimoto I, Kodama J, Seki N et al. Vascular endothelial growthfactor-C expression and its relationship to pelvic lymph nodestatus in invasive cervical cancer. Br J Cancer 2001; 85: 93–7.

10 Urabe A, Matsumoto T, Kimura M et al. Grading system oflymphatic invasion according to D2-40 immunostaining is usefulfor prediction of nodal metastasis in squamous cell carcinoma ofthe uterine cervix. Histopathology 2006; 49: 493–97.

11 Miyakuni Y, Matsumoto T, Arakawa A et al. Lymphatic invasionaccording to D2-40 immunostaining is a predictor of nodalmetastasis in endometrioid adenocarcinoma of the uterinecorpus. Pathol Int 2008; 58: 471–78.

12 Weidner N, Semple JP, Welch WR et al. Tumor angiogenesisand metastasis: Correlation in invasive breast carcinoma. NEngl J Med 1991; 324: 1–8.

13 Van der Auwera I, Cao Y, Pepper MS et al. First internationalconsensus on the methodology of lymphangiogenesis quantifi-cation in solid human tumours. Br J Cancer 2006; 95: 1611–25.

14 Gien LT, Covens A. Lymph node assessment in cervical cancer:Prognostic and therapeutic implications. J Surg Oncol 2009; 99:242–47.

15 Sironi S, Buda A, Picchio M et al. Lymph node metastasis inpatients with clinical early-stage cervical cancer: Detection withintegrated FDG PET/CT. Radiology 2006; 238: 272–79.

16 Hasumi K, Sakamoto A, Sugano H. Microinvasive carcinoma ofthe uterine cervix. Cancer 1980; 45: 928–31.

17 Roche WD, Norris HJ. Microinvasive carcinomaof the cervix:The significance of lymphatic invasion and confluent patterns ofstromal growth. Cancer 1975; 36: 180–86.

18 Leman MH, Benson WL, Kurman RJ et al. Microinvasive carci-noma of the cervix. Obstet Gynecol 1976; 48: 571–8.

19 Zivaljeviç M, Vujkov T, Ninèiç D et al. Microinvasive carcinomaof the cervix. Arch Oncol 2004; 12: 142–44.

20 Cimpean AM, Saptefrati L, Ceausu R, Raica M. Characteriza-tion of endoglin and Ki-67 expression in endothelial cells frombenign and malignant lesions of the uterine cervix. Pathol Int2009; 59: 695–700.

21 Hillen F, Griffioen AW. Tumour vascularization: Sprouting angio-genesis and beyond. Cancer Metastasis Rev 2007; 26: 489–502.

22 Wilting J, Kurtz H, OH S-J, Christ B. Angiogenesis and lym-phangiogenesis: Analagoues mechanism and homologousgrowth factors. In: Little CD, Mironov V, Sage EH, eds. VascularMorphogenesis: In Vivo, in Vitro, in Mente. Basel: BirkhauserVerlag, 1998; 21–34.



Detection of early lymphangiogenesis by lymphatic microvascular density and endothelial...

Documents

Transcript of Detection of early lymphangiogenesis by lymphatic microvascular density and endothelial...