697 p53 expression and gene mutation in operable non-small cell lung cancer (NSCLC)

178

Biology

tous pattern. Twenty-five adenocarcinoma (ADC) and 25 squamous cell carcinoma (SCC) were also studied. We investigated NCAM expression using immunohistochemistry on both frozen and formalin fixed tissues, with 123C3 antibody. Comparison was made with others neuroendccrine markers such as chromogranin A and synaptophysin, and electron microscopy was used as a gold standard. At least one malignant component expressed NCAM in 14/20 mesothelioma. We demonstrated a membranar expression of NCAM molecules in 10 cases, within the fusiform component of biphasic and sarcomatous mesothelioma. In 3 epithelial and 3 biphasic mesothelioma, epithelial components were positive in IO to 70% of the cells. Moreover, in 4 epihelial mesothelioma, 10 to 30% of stromal fibroblasts exhibited a strong focal positivity in the vicinity of tumor cells. Chromogranin expression was never seen, whereas synaptophysin was noticed in 3 cases beside NCAM expression. Neither ADC nor SCC were positive for these markers, We conclude that original expression of NCAM in mesothelioma is in keeping with that reported in other biphasic tumors. This has to be taken into account in the differential diagnosis between primary mesothelioma, and non small cell carcinoma (SCC, ADC and neuroendocrine carcinoma) metastatic to the pleura, and between mesothelioma and pleural sarcoma.

I

694

A combination study of P53 protein expression DNA ploidy on differential diagnosis between atypical metaplasia and intraepithelial cancer-spreading in early bronchial carcinoma

and

T. Ogata, Y. Inage, T. Yamamoto, E. Akaogi, H. Horiguchi, H. Kamma. lbaraki Prefectural Univ. of Health Sciences & Tsukuba Univ. Ibaraki, Japan It is important for surgical pathologists to differentiate intraepithelial cancerspreading from atypical metaplasia of the bronchial epithelium at histologic study of biopsies or surgical stumps of the bronchus. A combination study of p53 protein (P53) expression and DNA ploidy was done on the bronchial specimens in cases with early bronchial carcinoma limited to its bronchial extension. Twenty-six lesions of atypical metaplasia and 43 early bronchial carcinomas including 23 carcinomas in situ (CIS) or microinvasive carcinomas were examined by immunohistochemistry of P53 and image cytometry of DNA contents. Cancer cells of the primary lesion and its intraepithelial spreading areas showed similar P53 stain intensity and DNA ploidy. Therefore we made diagnosis as metaplasia, when atypical cells in the bronchial epithelium showed different P53 stain intensity and DNA ploidy from those of primary cancer lesions. Cancer cells were unexpectedly widespread to the adjacent bronchial epithelium even in early bronchial carcinoma. Nuclear P53 was overexpressed in some metaplastic lesions with severe atypia. DNA aneuploidy was only found in cancer cells. Frequency of P53 overexpression and DNA aneuploidy is as follows: P53 OE Nonal spithelium (17) SM without atypia (24) SM with mild atypia (14) SM with savers atypia (12) CISlmicroinvasiva carcinoma (23) SM: squamous metaplasia, 0; number of

DNA AP

0% 0% 0% 0% 76% lesions, OK; overexpression,

0% 0% 0% 25% 64%

AP; aneuploidy

The combination study may be practically useful to differentiate intraepithelial spreading of cancer cells from atypical squamous metaplasia in the bronchial specimens.

I

695

thickening nuclear margin. Recently we developed an image cytometry system using a color image analyzer to calculate geometric parameters (nuclear size, form factor, polymorphic factor and chromatin texture), and to estimate nuclear DNA ploidy and DNA synthetic phase fraction (SPF) simultaneously. We present here the correlation between the geometric parameters and DNA status (ploidy or SPF). Touch smear specimens from 36 non-small cell lung cancers (18 adenocarcinomas, 14 squamous cell carcinomas, two large cell carcinomas, and two carcinoids) were used for this study. We calculated geometric parameters and DNA contents of feulgenstained cancer cells by our image cytometry system. Thirty-six lung cancers were divided into two groups by a DNA index (DI = 1.3) (11 diploidy and 25 aneuploidy cases) and to two groups by a SPF value (SPF = 20%) (16 non-proliferative and 20 proliferative cases). DNA ploidy status was significantly correlated with three geometric parameters of nuclear size, form factor and polymorphic factor, while there was no difference in these geometric parameters between non-proliferative and proliferative groups of cancers. Furthermore, parameters of chromatin texture were significantly different between diploidy and aneuploidy groups of cancers, while there was no difference between non-proliferative and proliferative groups of cancers. In conclusion, morphologic abnormalities of cancer cells may be greatly influenced by DNA aneuploidy of the cells not but by their proliferative activity in NSCLC.

Nuclear abnormalities of cancer cells in morphology are greatly influenced by DNA ploidy but not by proliferative activity in non-small cell lung cancer (NSCLC)

T. Yamamoto, M. Noro, H. Horiguchi, H. Kamma, T. Ogata, M. Matsui, E. Akaogi, S. Ishikawa, K. Mitsui, T. Mitsui. Konan Hosp., Tsukuba Univ. & tbaraki Prefectural Univ. of Health Sciences, Ibaraki, Japan Nuclear abnormalities of cancer cells in morphology have been a diagnostic marker to differentiate cancer cells from normal cells. The abnormalities include enlarged nuclear size, polymorphism, chromatin aggregation, and

I

696

lmmunohistochemical detection of topoisomerase II a (Topolla) P-glycoprotein (Pgp), and multidrug resistance protein (MRP) in non small cell lung cancer (NSCLC) and in small cell lung cancer (SCLC) before and after treatment with topoll directed drugs

J. Kreisholt, M. Sorensen, M. Sehested. Laboratory Copenhagen, Denmark

P.B. Jensen, B.S. Nielsen, C.B. Andersen, and finsen Centres, Rigshospitalef, DK-2100

NSCLC and SCLC differ significantly in their clinical response to topolla directed drugs as etoposide and teniposide as NSCLC is virtually insensitive to single agent therapy, while SCLC responds in 2/3 of cases. Immunohistochemical analysis on paraffin-embedded tissue from 26 cases of untreated NSCLC, and 29 oases of untreated SCLC (10 with additional tumor biopsies after treatment with topoll drugs) gave the Results: NSCLC had significantly less topoll than SCLC as only 6126 NSCLC cases had r5% positive cells compared to 28/29 SCLC, and O/26 NSCLC had >25% positive cells compared to 26/29 SCLC. 7/10 SCLC cases showed a decrease in topoll after treatment. Pgp was detected in >5% of cells in only 3/26 NSCLC and in 6/29 SCLC, and MRP in 8/26 of NSCLC and 9/29 SCLC. After treatment there was an increase in MRP positivity in 4/10 and a decrease in l/10 cases of SCLC while an increase in Pgp was detected in 6/10 patients. In 2/10 cases both MRP and Pgp increased. In Conclusion: The major difference between untreated NSCLC and SCLC was in topoll alpha content. In treated SCLC, emergence of all 3 MDR phenotypes was detected.

0697

p53 expression and gene mutation in operable non-small cell lung cancer (NSCLC)

J. Niklinski, L. Chyczewski, M.A. Sipowicz’, M. Furman. Sialysfok Med. Univ, Balystok, Institute, Frederic, MD, USA

Y.-H. Shiao ‘, J. Laudanski, Poland, ’ National Cancer

We have examined p53 protein expression and gene mutations in 74 surgically treated NSCLC patients, and association of the p53 alterations with the clinicopathological and prognostic parameters. Mutations in exons 5 through 8 of the p53 gene were screened by single strand confirmation polymorphism (SSCP) and DNA sequencing. Accumulation of p53 protein was detected in paraffin-embedded tissue by immunostaining using MAb-DO7 (Dako). Overexpression of ~53, indicated by 210% of tumour cells positive after immunohistochemistry, occurred in 41 of 74 tumours (55%) whereas p53 mutations were found in 19 of 54 tumours (35%). The concordance rate between p53 expression and mutations was 75%. Gene mutations and/or accumulation of ~53 protein were common in both squamous cell (66%) and large cell (66%) carcinomas but significantly less

179

Biology

frequent in adenocarcinoma (37%; p = 0.01). Alterations in p53 were also associated with the presence of lymph node metastases (p = 0.02), but not with age, sex and stage. To ascertain a possible relationship between p53 status and prognosis, a Kaplan-Meier survival analysis was carried out. 8 of the 32 patients without p53 alterations died within the follow-up period, while 20 of 42 patients having p53 alterations died in the same period. Using a multivariate analysis p53 alterations were found to be significantly associated with poor prognosis (p = 0.01). We conclude that p53 alteration is an independent prognostic factor and the status of p53 is useful information for the management of patients with surgically treated NSCLC.

698 III

Adenocarcinomas of the lungs show a variable histology. We have subclassified such lesions into five cell types: hobnail, columnar, polygonal, mixed and goblet cell types, and investigated their relationship with K-ras mutations. Codon 12, 13 and 61 of the K-ras gene in 120 surgically resected pulmonaty adenocarcinomas were examined by the mutation-allele-specific amplification method. Point mutations were observed in 10% of the adenocarcinomas limited to K-ras codon 12 and the commonest base substitution (nine cases) was a G to T transversion. Of the five types, goblet cell lesions demonstrated the highest mutation index, which at 100% (6/6) was significantly different from that of all other cell types. No relationship between K-ras mutation and cigarette smoking or pathological stages was observed. From these findings, it appears that development of goblet-cell type adenocarcinomas of the lung may involve different carcinogenic mechanisms from adenocarcinomas of other cell subtypes.

Stereological estimates of nucleus/cytoplasm ratio and star volume on fiberoptic bi Jpsies are of prognostic value to discriminate survival in advanced squamous cell carcinoma of the lung

H.A. Carvalho, P.H.N. Saldiva, Paulo, Sri0 Pado, Brazil

T.Y. Takagaki,

V.1 Capelozzi.

PUBLISHED

data may help cell lung carciin histopatho-

ABSTRACTS

Pathology

High K-ras mutation rates in goblet-cell-type adenocarcinomas of the lungs

E. Tsuchiya’, R. Furuta”, N. Wada’. K. Nakagawa*, Y. Ishikawa2, B. Kawabuchi *, Y. Nakamura3, H. Sugano’. ’ Saitama Cancer Center, Tokyo; ‘Cancer Institute, Tokyo; 3 Univ. of Tokyo, Japan

I699

Conclusions: Our results indicate that histopathological to predict prognosis in patients with advanced squamous noma, and encourage the use of morphometric procedures logical analysis of this type of lung tumors.

Univ. of Go

Objective: The present work was designed to evaluate the role of association of morphometric and clinical parameters in contributing to establish prognosis of patients submitted to radiotherapy of advanced squamous cell carcinoma of the lung. Methods: Forty patients were included in this study. Clinical parameters such as sex, age, performance status, weight loss, stage, and histological grade were evaluated. Amount of necrosis, keratin, inflammatory infiltrate, vascularization, mitosis, starvolume of the nuclei, and nucleus/cytoplasm and nucleolus/nucleus ratios in the tumor were included as morphometric parameters. Morphometric studies were performed by means of point counting techniques. Group 1 patients (n = 22) were those with survival < 6 months; Group 2 (n = 10) was composed by patients with survival from 7 to 12 months and Group 3 (n = 8), included patients with survival > 12 months. To characterize these 3 groups of patients, models combining categorical and continuous variables were constructed, by means of discriminant analysis. Results: Weight loss, histologic grade, nucleus/cytoplasm ratio and star volume of the nuclei were selected during the backward procedure as relevant variables to characterize the 3 groups of patients. The overall sensitivity of the model was 90% (percentage of “grouped” cases correctly classified).

0700

Topoisomerase Ila (Top0 Ila) and mitotic cell count as proliferation markers in non-small cell lung cancer (NSCLC)

G.P.M. ten Velde, M. Raayman-Geutjes, F.B.J.M. Universit)! Maastricht Universitl: Maastricht, NL

Thunnissen.

Maastricht

Proliferation is a prognostic indicator in NSCLC and is usually determined with PCNA or Ki-67, or recently with the Ki-67 epitope MIB-I. Topo Ilcz is an iso-enzyme, detectable during late S-phase, peaks in G2M and is absent in GO and Gi. Aim: To compare the prognostic significance of Topo Ila! with MIB-1 and mitotic cell count in operable NSCLC patients. Patients and Methods: Of 143 NSCLC patients 97 were stage I, 31 stage II and 15 stage Illa. Routinely processed paraffin blocks were stained with an antibody against Topo IIO! and MIB-I and mitoses were counted. All three markers were expressed as number of positive cells per 1000 cells. Reproducibility was within 10% limits. Results: Topo lily labelling was observed in 131 and MIB-1 in 134cases. There was a high correlation between Topo II(Y and mitotic cell count (p < O.OOl), but no correlation between Topo Ila and MIB-1 (p = 0.61). Mitotic cell count was significantly related with a shorter survival (p = 0.003), Topo llu immunoreactivity borderline (p = 0.08) and MIB-1 not. Conclusion: Mitotic cell count is a strong index of proliferation and a good prognostic marker in NSCLC. Topo llol labelling is highly correlated with mitotic cell count, but it has borderline prognostic significance.

u

Metastatic patterns in lung cancer: Prognostic implications

G. Buccheri, D. Ferrigno. Hospital, Cuneo, ltaly

The Td pulmonary

division

of the ‘A. Car/e”

All previous studies have consistently reported that, in lung cancer, both clinical stage and each of the T, N, M categories are important prognostically (Eur Repir J 1994; 7: 1350). There is scarce information, however, concerning the value of each anatomical component making up a given T, N, or M category. In 653 consecutive patients with a newly diagnosed carcinoma of the bronchus, the exact anatomical involvement was recorded. Four hundred and forty-four of them were very well staged, receiving, at least, a computerized tomography of thorax, abdomen, and brain. Variables considered were 6 for the T category, 13 for N, and 8 for the M one. There were no missing values for any variable. Both univariate (Kaplan and Meyer estimates) and multivariate (Cox’s regression model) survival analyses were performed. Vmable

Beta

RR.

t-value

p-value

Brain metastases

0.9635 0.7643 0.7119 0.4315 0.3456 0.2792 0.4478 0.2793

2.6 2.1 2.0 1.5 1.4 13 1.5 1.3

5.4199 4.7735 4.0774 3.0403 2.5159 2.1729 2.1695 1.7959


prognostically

important

Bone metastases Liver metastases Lung metastases Mediastinal vessel invasion High paratracheal N2 Supraclavlcular N3 Pre- and retrotracheal N2

Cases included: 653; Chi* = 171.325, df = 27, p = 0.00000

Almost

all the variables

recorded

were found