4th LTBW Abstracts/Lung Cancer IO (1994) 347-373 355

tions (presumably as the result of repeated exposure to carcino- gens such as tobacco), and is at increased risk for developing multiple, independent morphological alterations and malignan- cies. The concept of field cancerization is equally applicable to the respiratory tract. Lung cancers may be associated with ex- tensive surface changes and secondary primaries are relatively frequent after both NSCLC and SCLC. The annual incidence of second malignancies after resections for Stage I NSCLC is 2-5%.

From our preliminary data, we can come to the following tentative conclusions: (1) There is a graded spectrum of mor- phological and genetic changes ranging from hyperplasia to in- vasive carcinoma; (2) widespread, focal lesions occur throughout the bronchial epithelium of cancer patients, helping to confirm the ‘field cancerization’ theory; (3) Increased prolif- eration rate and morphometric changes are early events, pre- sent in hyperplastic cells, while aneuploidy accompanies dysplasia, and ~53 over-expression is present in minor subsets of preneoplastic cells. The finding of widespread changes in- volving the bronchial, bronchiolar and alveolar epithelial sur- faces in association with both central and peripheral tumors confirms the ‘field cancerization’ theory

Topoisomerase IIa and @ gene expression in non-small cell lung cancer Giaccone G, van Ark-Otte J, Rubio G, Capranico G, Lopez R, Pinedo HM. Free University Hospital, Amsterdam, Netherlands.

Levels of expression of topoisomerase II, probably the (Y form, appear to be correlated to drug sensitivity in vitro. We in- vestigated the expression of the two top011 genes by Northern blotting and RNase protection on tumor samples of a total of 61 radically resected non-small cell lung cancer (NSCLC) patients. For about half of them normal lung tissue was also available. Thirty-one samples were of squamous cell lung cancer histological type. None of the patients received chemo- therapy. TopoIIol RNase protection proved to be more sensi- tive and allowed evaluation of more samples than Northern blotting. By RNase protection assay, topoIIa! expression was undetectable in only 12% of tumors, while 32,% of normal lung samples had low but detectable levels of expression. The levels of expression in tumors was in general much lower in NSCLC samples than in lung cancer cell lines. TopoIIfi expression was detected in most tumors and normal lung samples, by RNase protection assay. This finding was also confirmed by im- munohistochemistry, which revealed a similar distribution of topoIlfi staining in tumor and normal tissues. By Southern blotting, no rearrangement was observed in the topoIIcy gene. A semiquantitative PRC method is now being used for topoIIa gene expression in bronchoscopic biopsies of SCLC patients undergoing chemotherapy treatment.

~53 mutational spectrum in human cancers: clues to cancer etiol- ogy and molecular pathogenesis Harris CC, Gerwin BI, Lehman TA, Forrester K, Bennett WP. Laboratory of Human Carcinogenesis, NCI. NIH, Bethesda, MD 20892.

The ~53 tumor suppressor gene is well suited for analysis of mutational spectrum in human cancers: (a) most common gene-

tic lesion in human cancers; (b) reasonable size as a molecular target; and (c) may indicate selection of mutations with pathobiological significance. Mutations in the evolutionarily conserved codons of the ~53 tumor suppressor gene are com- mon in diverse types of human cancer [I]. The ~53 mutational spectrum differs among cancers of the colon, lung, esophagus, breast, liver, brain, reticuloendothelial tissues, and hemopoietic tissues. Analysis of these mutations can provide clues to the etiology of these diverse tumors and to the function of specific regions of ~53. Transitions predominate in colon, brain, and lymphoid malignancies. Mutational hotspots at CpG dinucleotides in codons 175,248, and 273 may reflect an endo- genous mutagenic mechanism, i.e. deamination of 5-methylcytosine to thymidine. The distribution of ~53 muta- tions at these codons with CpG dinucleotides also varies among different tissue sites. Oxy-radicals may enhance the rate of deamination. G:C to T:A transversions are the most frequent substitutions observed in cancers of the lung, breast, stomach and liver, and are more likely to be due to bulky carcinogen- DNA adducts. ~53 mutation and/or accumulation of post- translationally modified ~53 protein can be preinvasive events in bronchial, mammary or esophageal carcinogenesis [2,3]. Mutations at A:T base pairs are seen more frequently in esophageal than in other solid tumors. Most transversions in lung, breast, and esophageal carcinomas are dispersed among numerous conserved codons. In liver tumors in persons from geographic areas in which aflatoxin B, and hepatitis B virus are cancer risk factors, most mutations are at one nucleotide pair of codon 249 [4,5]. In geographic areas where hepatitis B and C viruses - but not aflatoxin B1 - are risk factors, the ~53 mutations are distributed in numerous codons. One hypo- thesis concerning generation of liver cancers with this mutation is: (a) aflatoxin Bl is metabolically activated to form the pro- mutagenic N7dG adduct; and (b) enhanced cell proliferation due to chronic active viral hepatitis allows both fixation of the G to T transversion in codon 249 of the ~53 gene and selective clonal expansion of the cells containing this mutant ~53 gene. The induction of skin carcinoma by ultraviolet light is indicated by the occurrence of ~53 mutations at dipyrimidine sites in- cluding CC to TT double base changes [6]. The ~53 mutational spectrum in radon-associated lung cancer from uranium miners also differs from lung cancer caused by tobacco smoking alone (21. In summary, these differences in mutational spectrum among human cancer types suggest: (a) the etiological con- tributions of both exogenous and endogenous factors to human carcinogenesis; (b) specific proliferative effects conferred by different mutant ~53 genes in different human cell types; and (c) hypotheses for investigation.

1 Hollstein M, Sidransky D, Vogelstein B, Harris CC. ~53 mutations in human cancers. Science 1991; 253: 49-53.

2 Vahakangas KH, Samet JM, Metcalf RA, Welsh JA, Bennett WP, Lane DP, Harris CC. Mutations of ~53 and ras genes in radon-associated lung cancer from uranium miners. Lancet 1992; 339: 576-580.

3 Bennett WP, Hollstein MC, Metcalf RA, Welsh JA, He A, Zhu S, Kusters 1, Resau JH, Trump BF, Lane DP, Harris CC. ~53 mutation and protein accumulation during multi

356 4th LTB W Abstracts/Lung Cancer 10 (1994) 347-373

stage human esophageal carcinogenesis. Cancer Res 1992; 52: 6092-6097.

4 Hsu IC, Metcalf RA, Sun T, Welsh J, Wang NJ, Harris CC. ~53 gene mutational hotspot in human hepatocellular carcinomas from Qidong, China. Nature 1991; 350: 427-428.

5 Bressac B, Kew M, Wands J, Ozturk M. Selective G to T mutations of ~53 gene in hepatocellular carcinoma from southern Africa. Nature 1991; 350: 429-431.

6 Brash DE, Rudolph JA, Simon JA, Lin A, McKenna GJ, Baden HP, Halperin AJ, Ponten J. A role for sunlight in skin cancer: UV-induced ~53 mutations in squamous cell carcinoma. Proc Natl Acad Sci USA 1991; 88: 10124-10128.

Insulin-like growth factors and their binding proteins in lung cancer Havemann K, Jaques G, Wegmann B, Noll K. Deparrment of Internal Medicine, Division of Hematology/Oncology, Baldingerstrasse. D-3550 Marburg, Germany.

Recent studies have shown that human lung cancer cell lines in tissue culture produce insulin-like growth factor-2 (IGF-2), contain high affinity receptors for IGF-I and IGF-2 and re- spond to exogenously added IGF-I and IGF-2 mainly via the IGF-1 receptor. Based on these findings we and others have proposed the concept that the IGFs may be important growth factors in lung cancer. Recently, it has been demonstrated, that a family of structurally related proteins which specifically bind the IGFs are involved in the modulation of IGF action. Six dif- ferent IGF binding proteins (designated IGFBP-I through IGFBP-6) have been purified and cloned. We have shown that SCLC and NSCLC in culture under serum-free conditions secrete a number of IGFBPs. These IGFBPs are differentially distributed between cell lines and primary tumors of small cell and non-small cell lung cancer cell lines. Treatment of SCLC cell lines with IGFs increased the DNA-synthesis significantly; this effect was reversible by the addition of recombinant IGFBP-2, but not by IGFBP-1.

The influence of various hormones and growth factors on the biosynthesis of the IGFs and their binding proteins was in- vestigated. While no change in the production of the IGFs was found, the expression pattern for IGFBPs was changed by the incubation with IGFs or retinoic acid (RA). With IGF-2 better than with IGF-I we could stimulate the expression of IGFBP-3 significantly, and could inhibit the expression of IGFBP-I, moderately. RA (I PM) on the other side induced an increase of all IGFBPs, mostly pronounced of IGFBP-3.

These results suggest that the IGF activity at the cellular level can be regulated by different secretion patterns of IGFBPs and may have important implications for understanding mechanisms by which IGFs and IGFBPs interact to regulate lung cancer cell growth.

The clinical implication of neuroendocrine markers in non-small cell lung cancer Hirsch FR, Senderovitz T. Department of Oncology, Bispebjerg Hospital, Copenhagen, Denmark.

For many years the non-small cell lung cancer has been con-

sidered as a chemoresistent type of tumor. However, within the last years several pre-clinical and some clinical studies have demonstrated that the distinction between small cell (SCLC) and non-small cell lung cancer (NSCLC) is not black and white. There seems to be a grey zone between those two main types of lung cancer, which might be characterized by the presence of neuroendocrine markers.

From several studies about 20-25% of all NSCLC tumors demonstrates neuroendocrine features with one or more of the following markers: neuron specific enolase (NSE), chromogranin, synaptophysin, Leu 7, neural cell adhesion molecules (NCAM), etc. The present overview will focus on the clinical implication of the presence of NE-markers in non- SCLC tumors.

Two published studies have looked on the clinical impact of the expression of NE-markers in surgically resected tumors. In a study by Berendsen et al. [l] patients with tumors having more than 50% positive NE-cells (MOC-I antibodies) had sig- nificant shorter survival than patients with NE-negative tumors after resection. Kibbelaar et al. ]2] analysed 308 surgically resected lung carcinomas of various histological subtypes, and NSCLC-tumors positive for Mab 123 C3 showed post- operative overall- and disease-free survival time significantly shorter than the 123 C3-negative tumors. The conclusion from these studies might be difficult to draw due to the retrospective character of the studies, relatively few patients included in the analyses and the different NE-markers used. More recently Sundaresan et al. [3] have published a preliminary report focus- ing on NE differentiation and survival in surgically treated cases of lung cancer and no correlation was found between NE- differentiation and survival, but the study suggested that NE- NSCLC was more highly metastatic than the non-NE-NSCLC.

Some chemotherapy studies have focused on the prognostic role of the presence of NE-markers. Graziano et al. [4] analysed 26 responders versus 26 non-responders and found that IO/26 (38%) responders were NE-positive compared to O/26 among the non-responders. Responders with two or more positive markers showed superior survival. In a study from our institu- tion [5] we analysed I I4 patients with inoperable adenocarcino- ma included in a prospective chemotherapy study and found a significantly higher response rate to chemotherapy in the NSE- positive group (44% vs. 17%). Although, we found a tendency to a longer survival for the NSE-group (median 262 days versus 159 days) there was no statistically significant. Van Zandwijk et al. [6] found a positive correlation between serum-NSE and response on chemotherapy in locally advanced or metastatic NSCLC, but no significant correlation to survival. Two preli- minary reports have recently been published: Carles et al. [7] examined 97 patients with NSCLC treated with chemotherapy. NSE was predictive for a better survival (median 11 months versus 7.4 months) compared to the NE-negative tumors, and in the study by Ruchdeschel et al. [8] including 237 resected NSCLC patients and 219 patients with extensive NSCLC or SCLC, NSE in patients with extensive disease was associated with an improved response to chemotherapy, but NE-markers had no influence on survival in resected patients.

From the studies in the literature it might be concluded that the presence of NE-markers are correlated to a higher response