determination ofproliferative activity in nasal polyps · friedenreich antigen (tf antigen), which...
TRANSCRIPT
7 Clin Pathol 1997;50:923-928
Determination of proliferative activity in nasalpolyps
Sergio Hassid, Marie-Pierre Degaute, Sandra Dawance, Katja Rombaut, Nathalie Nagy,Georges Choufani, Christine Decaestecker, Andre Danguy, Isabelle Salmon, Robert Kiss
Service d'Oto-Rhino-Laryngologie,CliniquesUniversitaires del'Hopital Erasme,BelgiumS HassidS DawanceG Choufani
Service d'AnatomiePathologiqueM-P DegauteK RombautN NagyI Salmon
Laboratoired'Histologie, Facultede Medecine,Universite Libre deBruxelles, Brussels,BelgiumC DecaesteckerA DanguyR Kiss
Correspondence to:Dr Kiss, Laboratoired'Histologie, Faculte deMedecine, Universite Librede Bruxelles, 808 route deLennik, 1070 Brussels,Belgium.
Accepted for publication2 September 1997
AbstractAims-To determine the level ofprolifera-tive activity in 39 nasal polyps with clearcut distinct clinical behaviour patterns.Methods-The 39 nasal polyps included 11polyps labelled as "single" and taken fromthe lateral nasal wall and the middleturbinate; 12 polyps labelled as "massive"and relating to diffuse polyposis involvingthe entire nasal cavity; six polyps labelledas "ASA" and relating to nasal polypsfrom patients with acetylsalicylic acidintolerance and asthma; and 10 polypsfrom cystic fibrosis related polyposis. Cellproliferation was determined by two inde-pendent methods: first, the computerassisted microscope analysis of isolatedFeulgen stained nuclei for themeasurement of the percentage of cells inthe S phase of the cell cycle; and second,the immunohistochemical evaluation of aproliferation associated protein by meansof the MIB 1 monoclonal antibody.Results-The cystic fibrosis related poly-posis exhibited the highest proliferativeactivity of all the clinically identified nasalpolyp groups. Acute inflammatory nasalpolyps exhibited a higher cell proliferationthan chronic ones. The results also showthat while the immunohistochemical de-termination of cell proliferation by meansof the MIB 1 monoclonal antibody is avaluable tool in determining cell prolif-eration in nasal polyps, the cytometricalimage analysis ofFeulgen stained nuclei isnot useful for this purpose.Conclusion-Cell proliferation activityidentifies cystic fibrosis as being distinctfrom the other nasal polyp groups.(7 Clin Pathol 1997;50:923-928)
Keywords: nasal polyps; proliferation; immunohisto-chemistry; Feulgen; image cytometry
Nasal polyposis is a common disease thatseverely affects daily life in most sufferers.' Forsome authors2 this disease constitutes a clinicalproblem rather than a histological one. Bern-stein et a13 state that in spite of recent advancesin basic science, particularly immunology andmolecular biology, the cause and pathogenesisof nasal and paranasal polyposis have not beenclarified. These authors report3 that, in theiropinion, although theories involving the dys-function of the autonomic nervous system ofthe nose4 as well as abnormalities in carbohy-drate metabolism' and vasomotor imbalance'
have been put forward, the two most plausibletheories involve allergy6 and inflamation.7 Inaddition to all these data, it should also beemphasised that, as stated by Coste et al,' res-piratory epithelium (and indeed any lining epi-thelium) is subject to a continuous turnoverthat requires epithelial cell proliferation anddifferentiation for cell renewal and mucosalrepair after various forms of aggression.8
Recently, we proposed a novel classificationof nasal polyps.' This classification relies onboth morphological criteria relating to mor-phonuclear features from isolated Feulgenstained nuclei (that is, quantitatively describingthe chromatin patterns) and glycohistochemi-cal characteristics from histological slidestreated with three lectins (peanut agglutinin(PNA isolated from Arachis hypogaea), wheatgerm agglutinin (WGA isolated from Triticumvulgaris), and gorse seed agglutinin (UEA-Iisolated from Ulex europaeus)) and one neogly-coconjugate carrying a Gal(f1-3)GalNAc gly-can residue, therefore related to the Thomsen-Friedenreich antigen (TF antigen), whichrepresents the acceptor sites of endogeneous-like PNAs." The data obtained from thesestudies showed that our morphologicalclassification of nasal polyps fitted in with theclassification of nasal polyps defined on thebasis of clinical criteria.9
In our present study, we refer to theabovementioned nasal polyp classification9 andinvestigate whether any significant differencesexist between the clinically related groups interms of proliferation. Proliferative activity wasdetermined by means of two independentmethods. The first involved the use of compu-ter assisted microscopy applied to Feulgenstained nuclei in order to quantify theproportion of cells present in the S phase of thecell cycle (S phase fraction, SPF index)." Thesecond method was an immunohistochemicalone which involved the use of the MIB 1 anti-body that recognises cell proliferation associ-ated human nuclear antigens (the Ki-67antigen) present in the late G,, S, G2, and Mphases of the cell cycle,'2 13 but absent in the G,and early G, phases. The MIB 1 monoclonalantibody reacts with recombinant parts of theKi-67 antigen (the 1002 base pair Ki-67 cDNAfragment). 13
MethodsHISTOPATHOLOGICAL DIAGNOSIS AND CLINICALDATA
Between 1993 and 1996 the medical recordsand samples ofnasal polyps were obtained for 39
923
on July 16, 2020 by guest. Protected by copyright.
http://jcp.bmj.com
/J C
lin Pathol: first published as 10.1136/jcp.50.11.923 on 1 N
ovember 1997. D
ownloaded from
Hassid, Degaute, Dawance, Rombaut, Nagy, Choufani, et al
patients. The diagnosis of polyposis was madewith a rigid endoscope. This method enabled usto differentiate between single polyps (the singlegroup; n = 11) found typically between thelateral nasal wall and the middle turbinate, anddiffuse polyposis (the massive group; n = 12),which involves the entire nasal cavity. The ASAgroup (n = 6) included nasal polyps frompatients with acetylsalicylic acid intoleranceestablished by clinical history. Provocation testswere not performed. The cystic fibrosis relatedpolyposis group (CF group; n = 10) includedpatients with well documented pathologies andpositive sweat tests.The histological slides of all the cases were
reviewed by the same pathologist on the basisof the criteria described by Hyams et al. 4 Thehistological characteristics in terms of acuteversus chronic inflammatory polyps were alsoreviewed in the light of the criteria described bythese authors.'4
SPECIMEN PREPARATIONSThe nasal polyp cases came from archivematerials (formalin fixed, paraffin wax embed-ded tissues). One paraffin wax block was avail-able for each case. Seven slides were cut fromeach block. The slides 1, 3, 4, 5, and 7 were5 ,um thick. The slides 1 and 7 were haema-toxylin and eosin stained for histological diag-nosis. The slides 3 and 5 were immunostainedby means of the anti-MIB 1 antibody (Immu-notech SA (Marseille, France), code no. 0505,clone MIB 1, dilution 1/50). Slide 4 served as anegative control (the primary antibody wasabsent) for slides 3 and 5. The procedure usedwas identical to the one detailed elsewhere.'5Slides 2 and 6 were 80 jim thick and were sub-
0.35 r
0.30 F
0.25 H
0.20 e
00.15
0.10 F
0.05
0.00
-0.05Single Massive ASA
Histoclinical groups
Figure 1 Development of the mean (small squares), standard error (large rectcstandard deviation (bars) values relating to the immunohistochemical MIB 1 lc
index in four nasal polyp groups. Single, polyps from the lateral nasal wall and tturbinate (n = 11). Massive, diffuse polyposis involving the entire nasal cavitiesASA, polyps from patients with acetylsalicylic acid intolerance (n = 6). CF, polipatients with cystic fibrosis (n = 10).
jected to a method that enables cell suspen-sions to be obtained (after pronase digestion).16They were cytocentrifuged on to glass slidesand submitted to the Feulgen reaction'6 todetermine the SPF index" (see below).
VARIABLE DETERMINATIONMIB 1 indexFor each of the 39 nasal polyps under study thepercentage of positively labelled MIB 1 cellswas recorded in 10 independent areas of thesurface (lining) epithelium on each of the twoslides available for any given case. Each areaincluded between 800 and 1200 cells at a x200magnification. The MIB 1 index representedthe number of positively labelled MIB 1 cellsdivided by the total number of cells counted(positive and negative).
SPF indexThe nuclear DNA content of each nasal polypunder study was assessed by means of a DNAhistogram computed on 300 Feulgen stainedcell nuclei. The computation was carried outby means of a SAMBA 2005 system (Alcatel-TITN, Grenoble, France) on the basis of aprocedure detailed elsewhere." 16 SAMBAsystem was equipped with a black and whiteCCD camera and a Leitz (Diaplan; Brussels,Belgium) microscope with a x1000 magnifica-tion (numerical aperture: 1.30).The proliferation index is the percentage of
cells engaged in the S phase of a cell cycle.Thus, the proliferation index corresponds tothe area under the curve between the GI and G2peaks in the DNA histogram. The proliferationindex was calculated on each DNA histogramin accordance with the procedure that we havedeveloped and which is detailed elsewhere."
DATA ANALYSISThe results are given in the figures as the meanwith standard deviations and standard errors.The statistical comparisons of the data wereperformed by means of either one way varianceanalysis (the Fisher F test) when the value dis-tribution was normal (checked by means of thex2 test) and the variances equal (checked by theBartlett test), or by the Mann-Whitney U testwhen the variances among groups analysedwere not equal. All the statistical analyses werecarried out using the Statistica/Dos software(Statsoft, Tulsa, Oklahoma, USA).
ResultsRELATION BETWEEN HISTOCLINICAL TYPE ANDPROLIFERATIVE ACTIVITYFigure 1 shows the cell proliferation level asassessed by means of MIB 1 immunohisto-chemistry on the four nasal polyp groups understudy.The cell proliferation index in the single
CF group was significantly lower than in the otherthree groups: p = 0.02 v massive; p = 0.01 v
ASA; p = 0.0002 v CF. The CF group
angles), and exhibited a significantly higher cell prolifera-ibelling tion index than the remaining three groups:the middle p = 0.04 v massive; p = 0.02 v ASA. The mas-
snfrom ) sive and ASA polyps exhibited a similar cellproliferation index (p > 0.05).
xa)-ac
.)
._
m
924
on July 16, 2020 by guest. Protected by copyright.
http://jcp.bmj.com
/J C
lin Pathol: first published as 10.1136/jcp.50.11.923 on 1 N
ovember 1997. D
ownloaded from
Proliferation in nasal polyposis
Li
Li
Massive ASAHistoclinical groups
Figure 2 shows that no statistically signifi-cant (p > 0.05) differences were found be-tween the various nasal polyp groups in termsof proliferative activity when this activity wasdetermined by means of the SPF index.An evident misfit is apparent between the
data obtained when figs 1 and 2 are compared.This misfit might be explained, at least in part,as follows. Figure 3A-C illustrates the variouspatterns ofMIB 1 immunostaining that can beobtained in nasal polyps. In fig 3A, only a fewepithelial cells are labelled positively by theMIB 1 monoclonal antibody. In fig 3B and C,numerous epithelial cells are labelled, but thelabelling pattern is distinct. In fig 3B, the label-ling pattern is homogenous. Indeed, the densityof the positive MIB 1 cells is homogenous inthe lining epithelium. In contrast, in fig 3C thedensity of the positive MIB 1 cells is hetero-genous, that is, there is a clonal growth pattern.
Figure 3D illustrates the type of biologicalmaterial on which the SPF index was deter-mined (the arrows indicate the cells that will be
CF retained in the final analysis).When the specimens shown in fig 3A-C are
compared, it seems that the distribution of thectangles), and proliferating cells in the lining epithelium isfraction by very heterogenous from one polyp to another.nasalpolyp This means that the architecture of the lining
epithelium must be retained if the proliferation
tX'>( \*rt1# # #
F
44*Sb
..aW
,, * z .
**X *ts 4,^' ~ ~~ ts' + *
S* 4#' . S>
.. o~~~
r..
a
.r
.t
r
*4
$#p;aXt:.:i14..!,*
* t'' *4*
,';" ..i ' iS '.
.:*.4S'*
. \lbOY.' 6S4
A,b
i.' 4%i5 '**..A*' *-{.
.,,,A * '5
.w
:-,P.4.
Figure 3 Illustration of cell proliferative activity as revealed by the MIB 1 monoclonal antibody in three nasal polyps. A nasal polyp whose surfaceepithelium is associated with a very low level ofproliferation (A). Two nasal polyps with a high level ofproliferative activity in the surface epitheliumshowing a diffuse pattern ofproliferation (B) and a clonal pattern ofproliferation (C). A Feulgen stained cytospin to be submitted to computer assisted
microscopy for determination of the S phase fraction (D). The arrows indicate those nuclei that will be retained in the final analysis.
2.2
1.8 H
1.4 _
1.0 H0
1/)U)
Li
0.6 _
0.2 H
-0.2 _
-0.6Single
Figure 2 Development of the mean (small squares), standard error (large re4standard deviation (bars) values relating to the determination of the S phase)means of the cytometrical image analysis of Feulgen stained nuclei in the fourgroups described in fig 1.
:
A:
B....rS,.*.. i
:: ;:
p* ..
..p9".;.
'A
;o -
4:t.,, ,
I'T.
.t
*
w
4.
.I.
ta. ,:: If .1
*'. tF ...
4*:!': i
925
*1 4
;c.. g...,.i.-O :":
ik #t.
*V-
on July 16, 2020 by guest. Protected by copyright.
http://jcp.bmj.com
/J C
lin Pathol: first published as 10.1136/jcp.50.11.923 on 1 N
ovember 1997. D
ownloaded from
Hassid, Degaute, Dawance, Rombaut, Nagy, Choufani, et al
CsA
.4
_~ t*¢
*~~~~~~ ,
ilP * 41.
4M6%, Jrv
Figure 4 Histological patterns (haematoxylin and eosin staining; original magnification, x200) of an acute inflammatorynasal polyp, characterised by the absence ofglands in the stroma (A), and two chronic inflammatory nasal polyps,characterised by the presence of either a few (B) or many (C) seromucinous glands in the stroma.
rate is to be determined correctly. Figure 3Dshows clearly that this architecture no longerexisted in the biological sample submitted foranalysis to the computer assisted microscope.
RELATION BETWEEN INFLAMMATORY STATE ANDPROLIFERATIVE ACTIVITYFigure 4A illustrates the histological pattern(on haematoxylin and eosin stained slides) ofan acute inflammation nasal polyp, character-ised by the absence of glands in the stroma.Two chronic inflammation nasal polyps, char-acterised by the presence of either a few (fig4B) or many (fig 4G) seromucinous glands inthe stroma are also shown.
Figure 5 shows that the pattern of prolifera-tive activity is significantly different (p = 0.02)
I _~~~~~~~Acute Chronic
Histoclinical groups
Figure 5 Development of the mean (small squares), standard error (large rectangles), andstandard deviation (bars) values relating to the MIB 1 immunohistochemical labellingindex in acute versus chronic inflammatory nasal polyps as described in fig 4.
in these two groups of acute versus chronicinflammatory nasal polyps.
DiscussionThe morphological appearance of the nasalpolyps on the 5 gm thick haematoxylin andeosin stained histological slides did not reflecttheir clinically observed behaviour. By way ofan example, Davidson and Hellquist' identifiedfour morphological types of nasal polyp: com-mon; chronic inflammatory; a type showinghyperplasia or seromucinous glands; and a typewith atypical stroma-a nasal polyp type thatcan be mistaken for a neoplasm.2 These fourmorphologically identified nasal polyp typesdid not fit in with our study of the four nasalpolyp groups selected on the basis of clinicalcriteria.9 Therefore, we attempted to establish anasal polyp classification on the basis ofmorphological criteria because this type ofclassification would be very helpful clinically,particularly in the monitoring of patients.9Because conventional morphology appears tobe insufficient to set up a classification of thistype, we turned to the more sophisticatedmethods of image cytometry and lectin glyco-histochemistry.Using the computer assisted microscope
analysis of Feulgen stained nuclei (thatenabled the morphometry of the cell nucleiand the chromatin patterns to be describedquantitatively by means of 26 variables), wewere able to identify four major groups ofnasal polyps: diffuse allergy related polyposis,cystic fibrosis related polyposis, single polypseither associated or unassociated with allergy,and a fourth group that included diffuse poly-posis not associated with allergy and ASArelated polyposis.'7 Thus, these four groups ofnasal polyps differed markedly in their morpho-nuclear characteristics, but the data obtainedfrom this first study took nuclear characteris-tics into consideration only. Therefore, wecompleted our investigation by a second studythat characterised the cytoplasmic and mem-brane related properties demonstrated by
0.40
0.35 _
0.30
0.25 _xa)
a,
0.
r-
0.20 H
0.15 _
0.10 _
0.05 H
0.00 _-
-11 nIh
926
on July 16, 2020 by guest. Protected by copyright.
http://jcp.bmj.com
/J C
lin Pathol: first published as 10.1136/jcp.50.11.923 on 1 N
ovember 1997. D
ownloaded from
Proliferation in nasal polyposis
means of lectin and neoglycoproteinhistochemistry.'° We observed that the group
of single polyps exhibited distinct glycohisto-chemical characteristics when compared withboth diffuse and cystic fibrosis related polypo-sis groups."0 It was by combining these twomethodological approaches (nuclear Feulgenstaining and cytoplasmic and membraneglycohistochemical staining) that we suc-
ceeded in setting up the abovementionedmorphology based classification for nasalpolyps.9The aim of the present study was, therefore,
to investigate whether the four groups of clini-cally identified nasal polyps exhibited distinctpatterns of proliferative activity.At the present moment, it remains unclear
how a polyp develops from normal nasalmucosa. As mentioned, of all the theories putforward, the predominate one concerns majordefects (probably induced by inflammatoryprocesses) that occur in the bioelectric integrityof the Na+ and Cl- channels on the luminalsurface of the epithelial respiratory cells.3Bernstein et al' argue that a change in Na+absorption may result in an increased move-
ment of water into a cell and the interstitialfluid with the ultimate result that all these dis-turbances lead to an oedema that may, in turn,lead to the growth and enlargement (by swell-ing) of the surface epithelium into a nasalpolyp. However, the theory that increased cellproliferation occurs in nasal polyps in compari-son with normal mucosa cannot be ignored.Indeed, as stated by Coste et al,' nasal polypsare not only characterised by the bulging of theoedematous process made up of a looseoedematous mucosal connective tissue, butalso by the covering of this connective tissue bythe respiratory epithelium that has an increasedsurface area compared with that of normalnasal mucosa.
Coste et al' made use of two independentmethods, flow cytometry and proliferation cellnuclear antigen (PCNA) immunohistochemis-try, to determine the proliferative activity in a
series of patients. For each patient, theycompared the epithelial cell proliferation innasal polyps and inferior turbinate nasalmucosa in which polyps never occur andreported that the percentage of cells in the Sphase was significantly higher (p < 0.05) in cellpopulations obtained from nasal polyps thancells from normal mucosa.' However, theyobserved some misfits between the flowcytometry and immunohistochemistry relatedresults. It should be remembered that althoughthe flow cytometry method allows the auto-matic quantification of the nuclear DNAcontent on the basis of cell cycle analysis of a
large number of individual cells (severalthousand) in a few minutes, it requires the dis-integration of the tissue to be analysed. Thismeans that the tissue architecture is no longerrespected and the so called epithelial compart-ment that, theoretically, is the only one
analysed by flow cytometry, may be contami-nated by many other cell types includinginflammatory and connective tissue cells. Inour present study, we made use oftwo methods
similar to those used by Coste et al': imagecytometry of Feulgen stained nuclei, which issimilar to flow cytometry and MIB 1 antigenimmunohistochemistry, which is also an im-munohistochemical method like PCNA. Ourdata show clearly that the disintegration of thetissue required for image cytometry analysisdid not enable us to demonstrate any specificproliferation pattern in the various nasal polypgroups while the MIB 1 immunohistochemicalmethod did enable this distinction to be made.This feature relies directly on the particulargrowth pattern of the surface epithelium innasal polyps, as shown in fig 3.Thus, our results show that among all the
clinically identified nasal polyp groups, it wasthe cystic fibrosis related ones that exhibitedthe highest proliferative activity. Therefore, thisbiological feature identifies cystic fibrosis asbeing clearly distinct from the other nasal polypgroups. In a recent editorial, Batsakis andEl-Naggar'5 wrote: "can sinonasal polyps frompatients with cystic fibrosis be distinguishedfrom atopic nasal polyps? Probably not". Itseems that the biological markers from themorphological field that we have detailedelsewhere,9 10 and those relating to the cell pro-liferative activity detailed here, could help indistinguishing between these two groups ofnasal polyps. The present study also shows thatacute inflammatory nasal polyps exhibit ahigher level of cell proliferation than chronicones.
This work was supported by grants awarded by the Fonds de laRecherche Scientifique Medicale (FRSM, Belgium).
1 Coste A, Rateau JG, Roudot-Thotaval F, Chapelin C, GilainL, Poron F, et al. Increased epithelial cell proliferation innasal polyps. Arch Otolaryngol Head Neck Surg1996;122:432-6.
2 Davidson A, Hellquist HB. The so-called "allergic" nasalpolyp. Otorhinolaryngology 1993;55:30-5.
3 Bernstein J, Gorfien J, Noble B. Role of allergy in nasalpolyposis: a review. Otolaryngol Head Neck Surg 1995;1 13:724-32.
4 Cauna N, Hinderer KH, Manzetti CW, Swanson EW Finestructure of nasal polyps. Ann Otol Rhinol Laryngol1972;81:41-58.
5 Slavin RG. Nasal polyps and sinusitis. In: Middleton E Jr,Reed CE, Ellis CE, Atkinson NF Jr, Yunginger JW, eds.Allergy: principles and practice, Vol 3, 3rd edn. St Louis: CVMosby, 1988:1291-303.
6 Holopainen E, Makinen J, Paavolainen M, Palva T, SaldoDR. Nasal polyposis: relationship to allergy and acetylsalicylic acid intolerance. Acta Otolaryngology 1979;87:330-4.
7 Small P, Frenkiel S, Black M. Multifactorial etiology ofnasal polyps. Ann Allergy 1981;46:317-20.
8 Breeze RG, Wheeldon EB. The cells of the pulmonary air-ways. Annu Rev Respir Dis 1977;116:705-777.
9 Hassid S, Decaestecker C, Hermans C, Salmon I, PasteelsJL, Danguy A, et al. Lectin glycohistochemical and Feulgencytometrical characteristics analyzed using an artificialintelligence-related algorithm to propose a new histoclini-cal classification for nasal polyposis. Ann Otol RhinolLaryngol. [In press.]
10 Hassid S, Salmon I, Brugmans M, Dawance S, Kiss R,Gabius HJ, et al. Histochemical study of epithelia of nasalpolyps by biotinylated lectins and neoglycoproteins. Acomparison with the normal respiratory epithelium. EurJMorphol 1997;35:77-85.
11 Salmon I, Kiss R. Relationship between proliferative activityand ploidy level in a series of 530 human brain tumorsincluding astrocytomas, meningiomas, schwannomas andmetastases. Hum Pathol 1993;24:329-35.
12 Gerdes J, Schwab U, Lemke H, Stein H. Production of amouse monoclonal antibody reactive with a human nuclearantigen associated with cell proliferation. Int J Cancer1983;31:13-20.
13 Cattoretti G, Becker MH, Key G, Duchrow M, Schulter C,Galle J, et al. Monoclonal antibodies against recombinantparts of the Ki-67 antigen (MIB 1 and MIB 3) detect pro-liferating cells in microwave-processed formalin-fixedparaffin sections. JPathol 1992;168:357-63.
927
on July 16, 2020 by guest. Protected by copyright.
http://jcp.bmj.com
/J C
lin Pathol: first published as 10.1136/jcp.50.11.923 on 1 N
ovember 1997. D
ownloaded from
Hassid, Degaute, Dawance, Rombaut, Nagy, Choufani, et al
14 Hyams VJ, Batsakis JG, Michaels L. Nasal and paranasalpolyposis. In: Hyams VJ, Batsakis JG, Michaels L, eds.Tumors ofthe upper respiratory tract and ear, 2nd edn. Atlas ofTumor Pathology. Washington DC: Armed Forces Instituteof Pathology, 1988;25:11 -33.
15 Kiss R, DeWitte 0, Decaestecker C, Camby I, Gordower L,Delbecque K, et aL The combined determination of prolif-erative activity and cell density to identify short- as opposedto long-surviving adult patients with supratentorial high-grade astrocytic tumors. Am Clin Pathol 1997;107:321-31.
16 Kiss R, Salmon I, Camby I, Gras S, Pasteels JL. Characteri-zation of factors in routine laboratory protocols which sig-nificantly influence the Feulgen reaction. .7 Histochemtz Cyto-chem 1993;41:935-45.
17 Hassid S, Choufani G, El-Kattabi 0, Dawance S, Decaes-tecker C, Brugmans M, et al. Image cytometry characteri-zation of ploidy level, proliferative activity and chromatinpattern in 50 nasal polyps. IntJ7 Oncol 1996;9:137-43.
18 Batsakis JG, El-Naggar AK. Pathology consultation. Cysticfibrosis and the sinonasal tract. Annii Otol Rhiiol Larvigol1996;105:329 30.
928
on July 16, 2020 by guest. Protected by copyright.
http://jcp.bmj.com
/J C
lin Pathol: first published as 10.1136/jcp.50.11.923 on 1 N
ovember 1997. D
ownloaded from