ccA

OnigSpptvMwgtcpRttvbmiCbpsts

LfWtcwphs

F

R

0

C

d

-myc and bcl-2 expression in supraglottic squamous cellarcinoma of the larynx

LI OZDEK, MD, SARP SARAC, MD, M. UMUT AKYOL, MD, ARZU SUNGUR, MD, and TANER YILMAZ, MD, Ankara, Turkey

foh

icbioc

ops

M

ltgMamstpsgrctt

C

6y

dgoe

vd

co

BJECTIVES: To evaluate the expression and prog-ostic significance of c-myc and bcl-2 oncogenes

n squamous cell carcinoma (SCC) of the supra-lottic larynx.TUDY DESIGN: A retrospective cohort study of 61atients who underwent surgery for SCC of the su-raglottic larynx. Gender, age, TNM status, opera-

ive procedure, recurrences, and disease-free sur-ival periods were recorded.ETHODS: Hematoxylin and eosin-stained sectionsere reexamined for grade, invasion of tumor mar-ins, lymphovascular invasion, lymphocyte infiltra-

ion, and perineural invasion. Immunohistochemi-al detection of c-myc and bcl-2 oncogenes waserformed using monoclonal antibodies.ESULTS: No correlation was observed between ei-

her c-myc or bcl-2 and the clinical and his-opathologic parameters. Survival analysis re-ealed no correlation of either c-myc (P � 0.88) orcl-2 (P � 0.85) with the disease-free survival. c-yc expression was found to be significantly higher

n bcl-2-positive patients (P � 0.001).ONCLUSION: Neither c-myc nor bcl-2 had shown toe prognostic factor for laryngeal carcinoma in thisresent study. Correlation between c-myc and bcl-2upports the experimental observations of coopera-ive action between these two genes in tumorigene-is. (Otolaryngol Head Neck Surg 2004;131:77-83.)

aryngeal squamous cell carcinoma (SCC) is the mostrequent tumor of the head and neck region in the

estern world. Improvement in either diagnostic orherapeutic modalities did not change therapeutic out-ome in last 30 years. TNM status, which is the mostidely used criteria in staging, is not sufficient toredict the tumor prognosis. Most recent studies onead and neck carcinoma have concentrated to under-tand tumor biology and to find more reliable predictive

rom the Department of Otolaryngology, Minister of Health, Ankara Training

and Research Hospital (Dr Ozdek) and the Department of Otolaryngol-

ogy (Drs Sarac, Akyol, and Yilmaz) and Department of Pathology (Dr

Sungur), Hacettepe University School of Medicine.

eprint requests: Ali Ozdek, MD, Inci bloklari mah, Ozgur-Anil sitesi A blok

30, 06530 100.Yl Ankara, Turkey; e-mail, Ozdek@ttnet.net.tr.

194-5998/$30.00

opyright © 2004 by the American Academy of Otolaryngology–Head and

Neck Surgery Foundation, Inc.

oi:10.1016/j.otohns.2004.02.015

actors. Recently, attention has been focused on cellularncogenes as potential prognostic predictors of theead and neck carcinoma.

The genes c-myc and bcl-2 are known to be involvedn cell cycle control and/or regulation of programmedell death and might have an impact on the biologicalehavior of tumors. However, most previous studiesnvestigating the clinical relevance of expression levelsf these genes in head and neck carcinoma yieldedonflicting results.

The aim of this study was to evaluate the expressionf c-myc and bcl-2 oncogenes and to investigate theirrognostic significance in squamous cell carcinoma ofupraglottic larynx.

ATERIALS AND METHODSFor this retrospective study, 61 patients were se-

ected from a total of 212 patients treated for SCC ofhe supraglottic larynx at the Department of Otolaryn-ology, Head and Neck Surgery, Hacettepe Universityedical Faculty, between 1975 and 1993. Following

re the exclusion criteria that were used to derive ourore homogenous study group: history of previous

urgery, radio- and/or chemotherapy, follow-up lesshan 3 years for patients without disease recurrence,reoperative tracheotomy, and positive margins in theurgical specimen. Patients’ charts were examined andender, age, TNM status, surgical procedure, diseaseecurrence, and disease-free survival periods were re-orded. Pathological examinations included conven-ional histopathology and immunohistochemical iden-ification of c-myc and bcl-2.

onventional HistopathologyFor conventional histopathology, we reexamined

-�m paraffin sections that were stained with hematox-lin and eosin.

Grading. Tumors were graded as grade 1 (wellifferentiated), grade 2 (moderately differentiated), orrade 3 (poorly or undifferentiated) based on the degreef keratinization, nuclear pleomorphism, and the pres-nce or absence of intracellular bridges.

Invasion pattern of the tumor margin. In-asion pattern of tumor is classified as “pushing bor-ers” and “infiltrating borders.”

Lymphovascular invasion. Each case waslassified as positive or negative based on the presencer absence of tumor cells in the lymphovascular spaces.

77

tgc

pt

I

bmnawtdpiaasmbatPuaaiikcewTqsTo2Ts

S

fWatuee

RP

wTT2tgwpNocsuntsthsm

C

2Igiswti(

c

6(“i“(pa

S

blgn

Otolaryngology–Head and Neck Surgery

78 OZDEK et al July 2004

Lymphocyte infiltration. Lymphocytic infiltra-ion of the advancing tumor border was evaluated andraded as poor, moderate, or marked using arbitraryriteria.

Perineural invasion. Cases were classified asositive or negative based on the presence or absence ofumor cells in the perineural spaces.

mmunohistochemistryFor immunohistochemical detection of c-myc and

cl-2, 6-�m sections taken from paraffin blocks wereounted on glass slides and incubated at 56° C over-

ight. The slides were then deparaffinized by xylene,nd dehydrated with alcohol. After washing with tapater, the slides then were incubated in 80% methanol

hat contained 0.3% hydrogen peroxide, to block en-ogenous peroxidase activity. After washing with phos-hate-buffered saline (PBS) solution, the slides weremmersed in a pH 6 citrate buffer solution and placed in

microwave at 750 W for two 5-minute periods tochieve antigenic retrieval. After washing with PBSolution, the slides were incubated with monoclonalouse anti-human c-myc and bcl-2 oncoprotein anti-

odies (DAKO, Glostrup, Denmark) diluted with PBSt 1:100 and 1:40, respectively, for 2 hours at roomemperature. Next, the slides were again washed inBS. Labeling and visualization were performed bysing biotin-conjugated anti-mouse serum (DAKO)nd 0.1% diaminobenzidine stain. Appropriate positivend negative controls were prepared in the same fash-on. Our coauthor (A.S.), who is a pathologist, exam-ned the slides under the light microscope withoutnowledge of any case histories. Nuclear and perinu-lear staining were accepted as positive for c-myc pres-nce (Fig 1), cytoplasmic and membranous stainingere accepted as positive for bcl-2 presence (Fig 2).he frequency of positive cells was evaluated semi-uantitatively and determined as the ratio of positivetained tumoral cells to overall tumoral cell population.hey were graded from 0 to �3 as follows: 0 � nonef the tumoral cells, �1 � �25% of tumoral cells, �2 �5-50% of tumoral cells, �3 � �50% of tumoral cells.he intensity of staining was not graded and any expres-ion of c-myc and bcl-2 were classified as positive.

tatistical AnalysisStatistical analysis was carried out using the SPSS

or Windows version 9.0 computer program. Kruskal-allis test was used to investigate the relation of c-myc

nd bcl-2 with clinical and histopathological parame-ers. Correlation analysis with the Pearson method wassed to detect any correlation between c-myc and bcl-2xpression. The log rank test was used to analyze theffects of c-myc and bcl-2 on disease-free survival.

ESULTSatients

All 61 patients were males of 34 to 68 years of age,ith a mean age of 53 years. Thirteen individuals had1, 39 had T2, 5 had T3, and 4 had T4 lesions.hirty-two patients had N0, 14 had N1, 13 had N2, andhad N3 neck at their initial clinical evaluation. Thirty-

hree patients had undergone partial and 28 had under-one total laryngectomy. Unilateral neck dissectionas performed in 19 patients because our policy was toerform neck dissection only to the ipsilateral site inO neck without intraoperative frozen section analysisf the lymph nodes until 1985. After the year 1985, wehanged our policy and performed bilateral neck dis-ections in all supraglottic tumors and 42 patients hadndergone bilateral neck dissection. Cervical lymphode metastasis was present in 33 (54.1%) patients. Ofhese, 9 patients (27.3%) exhibited extracapsular inva-ion. Locoregional recurrence was observed in 7 pa-ients, and distant metastasis in 5 cases. Two patientsad both locoregional recurrence and distant metasta-es. Disease-free patient survival ranged from 3 to 100onths, with a mean of 50 months.

onventional HistopathologyConventional histopathologic examination revealed

6 patients (42.6%) with grade I, 25 (41.0%) with gradeI, and 10 (16.4%) with grade III tumors. Tumor mar-ins were found to be pushing in 47 (77 %) andnfiltrating in 14 (23%) patients. Lymphovascular inva-ion was positive in 16 (26.2%) and perineural invasionas positive in 7 (11.5%) patients. Lymphocytic infil-

ration of the advancing tumor border was labeled poorn 3 (4.9%), moderate in 37 (60.7%), and marked in 2134.4%) cases.

-myc and bcl-2 Immunohistochemistryc-myc expression was observed in 56 (91.8%) of the

1 patients. Of these, 23 (41.1%) had “1�”(mild), 2035.7%) had “2�”(moderate), and 13 (23.2%) had3�” (intense) staining. bcl-2 expression was observedn 22 (36.1%) of 61 patients. Of these, 14 (23.0%) had1�”(mild), 6 (9.8%) had “2�”(moderate), and 23.3%) had “3�” (intense) staining. The distribution ofatients according to the staining intensity for c-mycnd bcl-2 expression was given in Figure 3.

tatistical AnalysisKruskal-Wallis test failed to show any correlation

etween c-myc or bcl-2 and the clinical and histopatho-ogic parameters, namely, T stage, N stage, tumorrade, tumor margins, lymphovascular invasion, peri-eural invasion, lymphocyte infiltration, cervical lymph

Fna

Otolaryngology–Head and Neck SurgeryVolume 131 Number 1 OZDEK et al 79

ig 1. A) Positive nuclear staining of c-myc oncoprotein antibodies in the poorly differentiated squamous cells (immu-operoxidase �460). B) Corresponding H&E (�230) section of poorly differentiated squamous cell carcinoma formed bynaplastic cells with abundant mitosis.

FCt

Otolaryngology–Head and Neck Surgery

80 OZDEK et al July 2004

ig 2. A) bcl-2 staining is present in the moderately well differentiated tumor cells (immunoperoxidase �115). B)orresponding H&E (�230) section of moderately well-differentiated squamous cell carcinoma with island of tumor cells

hat exhibits a small amount of keratin.

nmi�(btt

D

cictmt

ofwhbs

cScScpcnctshc

efcip

rwhewpso

esbnab

Tcp

Otolaryngology–Head and Neck SurgeryVolume 131 Number 1 OZDEK et al 81

ode metastasis, locoregional recurrences and distantetastasis (Table 1). Kaplan-Meier survival analysis

ndicated that there was no correlation between bcl-2 (P0.85) or c-myc (P � 0.88) and disease-free survival

Figs 4 and 5). A significant correlation was observedetween c-myc and bcl-2 expression (Pearson correla-ion � 0.42, P � 0.001). c-myc expression was foundo be significantly higher in bcl-2-positive patients.

ISCUSSIONIdentification of a malignancy’s aggressive biologi-

al behavior and/or the tendency to recur or metastasizes of paramount importance to the physician wherehoice of treatment modality is concerned. Thus, manyumor markers, cellular oncogenes, and proliferationarkers1 are currently under extensive study as poten-

ial prognostic indicators of head and neck SCC.Recently, attention has been focused on cellular

ncogenes. One such oncogene is c-myc, which codesor a nuclear regulatory protein that has been associatedith tumorigenesis.2 Amplification of the c-myc geneas been linked to poor prognosis in neuroblastoma,reast, gastric, colorectal, cervical, endometrial andmall-cell lung cancers.2

Overexpression of c-myc has been labeled an indi-ator of poor prognosis in nasopharyngeal carcinoma.3

till, these same authors found no correlation between-myc expression and clinicopathologic parameters.imilarly, Miyazaki et al4 found no correlation between-myc expression and clinicopathologic parameters andrognosis in esophageal carcinoma. However, in-reased propensity for lymph node metastasis has beenoted in patients with high-level c-myc expression inases of head and neck SCC.5 These data conflict withhose reported by Gapany et al,2 who found the aggres-ive behavior and increased metastatic propensity of theead and neck tumors to be coincidental with low-myc expression. In general, the data related to c-myc

Fig 3. Distribution of c-myc

xpression in laryngeal carcinoma are very limited. Soar, it has been shown that resistance of human larynxarcinoma cells to treatment with cisplatin, gammarradiation, and methotrexate does not involve overex-ression of c-myc or c-Ki-ras oncogenes.6

A large range of c-myc expression rates has beeneported in head and neck tumors in previous studies,hich was as low as 10% as in Takes’ study7 and asigh as 100% as in Sakai’s study.8 We found c-mycxpression in 91.8% of patients in the present study,hich was a relatively high rate. In 41.1% of theseatients, however, very mild degree of c-myc expres-ion was detected. Only 33 of 61 patients showed mildr intense degree of c-myc expression.

Another oncogene that was related with tumorigen-sis is the bcl-2 gene, whose product prolongs cellurvival. It was shown that overexpression of bcl-2 haseen related with poor outcome in early stage head andeck cancers.9 Studies of this gene in cases of esoph-geal carcinoma have revealed conflicting results. Sar-ia et al10 and Patel et al11 reported no correlation

cl-2 expression in patients.

able 1. The results of statistical analysisomparing the clinical and histopathologicalarameters with c-myc and bcl-2 expression

P-value

c-myc bcl-2

T stage 0.14 0.57

N stage 0.42 0.45

Tumor grade 0.29 0.75

Tumor margins 0.81 0.81

Lymphovascular invasion 0.39 0.52

Perineural invasion 0.08 0.60

Lymphocyte infiltration 0.06 0.58

Cervical LN metastasis 0.27 0.13

Locoregional recurrences and distant

metastasis

0.89 0.81

and b

bricdhKi

cTipwl

k

e pat

Otolaryngology–Head and Neck Surgery

82 OZDEK et al July 2004

etween bcl-2 expression and other histopathologic pa-ameters and survival. In contrast, Ohbu et al12 foundncreased bcl-2 expression to be correlated with a de-reased incidence of lymph node metastasis and longerisease-free survival. Reports on laryngeal carcinomaave revealed no correlation of bcl-2 with survival.13,14

latka et al15 found that the expression of bcl-2 proteinn T lymphocytes from the patients with laryngeal

Fig 4. Kaplan-Meier survival analysis of th

Fig 5. Kaplan-Meier survival analysis of th

arcinoma was significantly higher than in the controls.hey also they proposed that bcl-2 protein may interact

n the regulation of apoptosis of lymphocytes, takingart in anti-cancer defense. In our study, no correlationas found between bcl-2 and prognosis or histopatho-

ogic parameters.Tumorigenesis is a complex process with many

nown and unknown contributors. Synergistic action of

ients stratified by bcl-2 staining intensity.

ents stratified by c-myc staining intensity.

e pati

mcbdfatstvpeashgscstplPp

C

tpc

R

1

1

1

1

1

1

1

1

1

1

2

2

Otolaryngology–Head and Neck SurgeryVolume 131 Number 1 OZDEK et al 83

ultiple oncogenes may exist in this process. p53/bcl-2oexpression in laryngeal carcinoma was shown that toe related with shortened survival by Jackel et al16 Thisata conflict with findings reported by Hotz,17 whoound no correlation between p53 and bcl-2 in locallydvanced head and neck cancer. Fanidi et al18 showedhat bcl-2 specifically abrogates c-myc induced apopto-is without affecting the c-myc mitogenic function inheir experimental study. They proposed that this pro-ides a novel mechanism for oncogene cooperation ofotential importance both in carcinogenesis and in thevolution of drug resistance in tumors. Bissonnette etl19 found similar findings to Fanidi’s study. Suchynergistic actions of c-myc and bcl-2 in carcinogenesisave been described for basaloid SCC of the esopha-us20 and medullary thyroid carcinoma.21 In the presenttudy, a significant correlation was observed between-myc and bcl-2 expression. However it is difficult totate that c-myc and bcl-2 act as an interactive pair inhe carcinogenesis of laryngeal carcinoma just by de-ending on this finding. Investigations involving mucharger series and other well-known oncogenes, such asCNA, p53, nm23, cyclin d1, EGF, RB, are needed torove such an interaction.

ONCLUSIONIn conclusion, neither c-myc nor bcl-2 were shown

o be prognostic factors for laryngeal carcinoma in thisresent study, but coexpression of these two genes mayontribute to carcinogenesis.

EFERENCES1. Sarac S, Ayhan A, Hosal AS, et al. Prognostic significance of

PCNA expression in laryngeal cancer. Arch Otolaryngol HeadNeck Surg 1998;124:1321-4.

2. Gapany M, Pavelic ZP, Kelley DJ, et al. Immunohistochemicaldetection of c-myc protein in head and neck tumors. Arch Oto-laryngol Head Neck Surg 1994;120:255-9.

3. Porter MJ, Field JK, Leung SF, et al. The detection of the c-mycand Ras oncogenes in nasopharyngeal carcinoma by immunohis-tochemistry. Acta Otolaryngol 1994;114:105-9.

4. Miyazaki S, Sasno H, Shiga K, et al. Analysis of c-myconcogene in human esophageal carcinoma: immunohisto-chemistry, in situ hybridization and Northern and Southern

blot studies. Anticancer Res 1992;12:1747-56.

5. Haughey BH, Von Hoff DD, Windle BE, et al. c-myc oncogenecopy number in squamous carcinoma of the head and neck. Am JOtolaryngol 1992;13:168-71.

6. Osmak M, Beketic-Oreskovic L, Matulic M, et al. Resistance ofhuman larynx carcinoma cells to cisplatin, gamma-irradiationand methotrexate does not involve overexpression of c-myc orc-Ki-ras oncogenes. Mutat Res 1993;303:113-20.

7. Takes RP, Baatenburg de Jong RJ, Schuuring E, et al. Markersfor assessment of nodal metastasis in laryngeal carcinoma. ArchOtolaryngol Head Neck Surg 1997;123:412-9.

8. Sakai H, Kawano K, Okamura K, et al. Immunohistochemicallocalization of c-myc oncogene product and EGF receptor in oralsquamous cell carcinoma. J Oral Pathol Med 1990;19:1-4.

9. Friedman M, Grey P, Venkatesan TK, et al. Prognostic significanceof bcl-2 expression in localized squamous cell carcinoma of thehead and neck. Ann Otol Rhinol Laryngol 1997;106:445-50.

0. Sarbia M, Bittinger F, Porschen R, et al. Bcl-2 expression andprognosis in squamous cell carcinomas of the esophagus. Int JCancer 1996;69:324-8.

1. Patel DD, Bhatavdekar JM, Chikhlikar PR, et al. Clinical signif-icance of p53, nm23, and bcl-2 in T3-4 N1M0 oesophagealcarcinoma: an immunohistochemical approach. J Surg Oncol1997;65:111-6.

2. Ohbu M, Saegusa M, Kobayashi N, et al. Expression of bcl-2protein in esophageal squamous cell carcinomas and its associ-ation with lymph node metastasis. Cancer 1997;79:1287-93.

3. Spafford MF, Koeppe J, Pan Z, et al. Correlation of tumormarkers p53, bcl-2, CD34, Cd44H, CD44v6 and Ki-67 withsurvival and metastasis in laryngeal squamous cell carcinoma.Arch Otolaryngol Head Neck Surg 1996;122:627-32.

4. Whisler LC, Wood NB, Caldarelli DD, et al. Regulators ofproliferation and apoptosis in carcinoma of the larynx. Laryngo-scope 1998;108:630-8.

5. Klatka J, Rolinski J, Kupisz K, et al. Expression of bcl-2protein in lymphocytes of patients with laryngeal carcinoma.Eur Arch Otorhinolaryngol 1999;256:299-302.

6. Jackel MC, Sellmann L, Dorudian MA, et al. Prognostic signif-icance of p53/bcl-2 co-expression in patients with laryngealsquamous cell carcinoma. Laryngoscope 2000;110:1339-45.

7. Hotz MA, Bosq J, Zbaeren P, et al. Spontaneous apoptosis andthe expression of p-53 and bcl-2 family proteins in locallyadvanced head and neck cancer. Arch Otolaryngol Head NeckSurg 1999;125:417-22.

8. Fanidi A, Harrington EA, Evan GI. Cooperative interaction be-tween c-myc and bcl-2 proto-oncogenes. Nature 1992;359:554-6.

9. Bissonnette RP, Echeverri F, Mahboubi A, et al. Apoptotic celldeath induced by c-myc is inhibited by bcl-2. Nature 1992;359:552-4.

0. Sarbia M, Loberg C, Wolter M, et al. Expression of bcl-2 andamplification of c-myc are frequent in basaloid squamous cellcarcinomas of the esophagus. Am J Pathol 1999;155:1027-32.

1. Wang DG, Liu WH, Johnston CF, et al. Bcl-2 and c-Myc, but notbax and p53, are expressed during human medullary thyroid

tumorigenesis. Am J Pathol 1998;152:1407-11.