Acta Oncologica

ISSN: 0284-186X (Print) 1651-226X (Online) Journal homepage: http://www.tandfonline.com/loi/ionc20

Abstracts of Theses from the Scandinavian Countries To cite this article: (1992) Abstracts of Theses from the Scandinavian Countries, Acta Oncologica, 31:8, 867-875, DOI: 10.3109/02841869209089721 To link to this article: http://dx.doi.org/10.3109/02841869209089721

Published online: 08 Jul 2009.

Submit your article to this journal

Article views: 8

View related articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ionc20 Download by: [37.44.207.87]

Date: 21 January 2017, At: 04:50

Acrri Oncologicir Vol. 31. No. 8, pp. 867- 875. 1992

Abstracts of Theses from the Scandinavian Countries Ahstrncts of' Scandinuviun [hews on oncologic subjecrs ure puhlisheri under this lieuding. The full theses are us (I rule published hy (he univer.sities or us supplements to different journals. They ciin usually he ohtirined after contnct ivirh the aurhor.

Natural history, palliative treatment and prognostic factors in prostatic cancer- A population-based study from Orebro, Sweden. J.-E. JOHANSSON Department of Surgery, Cancer Epidemiology Unit and Division of Urology, University Hospital, S-751 85 Uppsala. Sweden. The aims of this study were to investigate the natural course, progression, disease-specific survival and overall survival in patients with untreated, localised prostatic cancer; and t o investigate, by means of two randomized trials, the safety and efficacy of different palliative treatments in advanced disease. All studies were population-based and prognostic factors were evaluated in both the localised and advanced stages. Patients (n = 223) with localised disease who received no initial treatment had a high (86.8%) 10-year corrected survival rate. Fifty-eight of them who met our current indications for radical prostatectomy had similar (87.9%) survival rates. The risk of a progressive and eventually fatal prostatic cancer decreased after five years of observation unlike the risk of dying of other Causes. which exceeded by far the mortality due to prostatic cancer. Only a few patients experienced more substatial symptoms because of local tumor growth and these were in general successfully treated hormonally. There seems to be a group of patients having 'latent' cancer with some local tumor growth but no metastatic potential. There is little room for therapeutic improvement with radical treatment (surgery or irradiation) at the time of diagnosis and such treatment should be evaluated in randomized trials with an untreated control group. In advanced stages. estrogen treatment (74 patients) resulted in a significantly ( p = 0.05) longer progression-free survival than did orchiectoniy (76 patients). However, because of a higher rate of cardiovascular complications. estrogen cannot be recommended in the combined oral and parenteral form used in this study. In therapy-resistant cases the response was better with medroxyprogesterone acetate ( 5 1 patients) than with estramustine ( 5 1 patients). The prognosis was poor, however, the mortality rate being 70% during the first year. The only significant prognostic factor for disease-specific survival in localised disease was tumor grade. In advanced disease. M category, T category. sedimentation rate and age were all related to both progression and to disease-specific death. Among 300 consecutive cases in all stages, the prognostic factors of significance for disease-specific surival were M category, T category, grade. performance status, sedimentation rate. and hemoglobin. The relationship of the sedimentation rate and prognosis was non-linear. The risk was lowest with a sedimentation rate of 40-50 mm/h and higher for both lower ( < 10 mm/h) and higher values ( > 6 3 mm/h). Sedimentation rate may be a marker in patients with a weak host defence or in those with proliferative disease. November 1991

Gastric stump carcinoma-A clinical study on carcinoma in the gastric remnant after surgery for benign gastroduodenal disease C. STAELVON HOLSTEIN Department of Surgery, Lund University, S-221 85 Lund, Sweden Mortality was studied after partial gastrectomy in a cohort of 1575 patients operated on due t o benign gastroduodenal di 29 to 59 years ago. Overall mortality was significantly higher than in the general population due to malignancies, respiratory disease or suicide. Patients operated on at young age constitute a definite risk group with an increased mortality due to smoking-related diseases and suicide during the first 19 years postoperatively. and an increased risk of death due to gastrointestinal malignancy, diseases and cancers in the respiratory organs. or suicide during the following years. In an attempt to reduce mortality in gastric stump carcinoma a defined cohort of operated peptic ulcer patients was followed up with an endoscopic screening programme during 17 years. No difference in gastric cancer mortality was found between patients subjected to screening compared to those who were not investigated. Among unselected patients who have undergone partial gastrectomy, endoscopic screening for gastric stump carcinoma is not worthwhile. The incidence of stump cancer, the compliance of patients, and the sensitivity of our screening instrument are all too low, and the risks in eldery patients not negligible. Patient history and endoscopic findings are poor instruments with which to detect patients with dysplasia or malignancy. Dysplasia in the gastric remnant can only be found with multiple biopsies. preferably from the anastomotic region. Patients with moderate and, especially. severe dysplasia in the gastric remnant have a high risk of gastric malignancy. Severe dysplasia calls for close endoscopic surveillance at short intervals whereas investigation biannually seems sufficient for patients with moderate dysplasia. Twenty-three patients with early gastric stump carcinoma were subjected to re-resection of the anastornotic area with a resection margin of 2 cm t o the tumour. Six patients developed recurrence and three died from gastric carcinoma during the follow-up period. It is concluded that total gastrectomy is the treatment of choice for early gastric stump carcinoma. Dietary intake, triolein breath test, anthropometry and biochemical nutritional status were compared in two groups of patients after total gastrectomy ( n = 10) and after partial gastrectomy ( n = 10). N o substantial nutritional benefits could be found with a small gastric remnant linked to a Roux loop, as compared to a Roux loop after total gastrectomy. March 1992

Cytogenetic and molecular genetic analyses of human malignant pleural mesothelioma M. TIAINEN Department of Medical Genetics, University of Helsinki, Helsinki. Finland Mesothelioma is an asbestos-related rare neoplasm. the cytogenetic and molecular features of which are not well known. We performed successful cytogenetic analyses on tumor and pleural effusion samples of 34 patients with ditl'use malignant pleural mesothelioma, 67% of whom had been exposed to asbestos. I n eight of these, cells with normal karyotype and/or cells with nonclonal abnormalities were detected. Clonal chromosomal changes were found in 26 patients. most of them having a

867

868

ABSTRACTS OF THESES

near-diploid stemline. Most of the karyotypes were complex, involving several structural and numerical chromosomal aberrations, and no tumor-specific abnormality was found. Frequently observed polyploidization and heterogeneity of various aberrations in individual cases were indicative of clonal evolution. The most frequent chromosomal losses occurred in chromosomes 1 (overlapping area of losses in Ipl l-p36.3), 3 (overlapping area of losses in 3p13-p21), and 22 which present candidate chromosomes carrying tumor suppressor genes. On the other hand, chromosomal gain may be a mechanism evaluating the expression of oncogenes o r growth factor genes. The most frequent chromosomal gains observed were total or partial polysomy of chromosomes 1 (q-arm), 7, and 11. Deletions and unbalanced translocations were the most frequent types of structural chromosomal aberrations. Clustering of breakpoints was detected in IplI-p22, lq12-q25, 3pl I-p21, and 7ql1.2. Similar patterns of chromosomal aberrations have been detected in 92 mesotheliomas published by other investigators. The general complexity of chromosomal changes in mesothelioma reflect the fact that the tumors analyzed have progressed into late stages in their evolution. Clearly most of the chromosomal aberrations are secondary and related to tumor progression. Aberrations of chromosomes I , 7 and 22 were further studied at the DNA level by RFLP analysis. No gene amplifications or rearrangements were found in NRAS, EGFR, MET, EPO, PGY3, PLANHI, PDGFA, and PDGFB in 23 cytogenetically characterized mesothelioma tumors. In two of the tumors, numerical aberrations of chromosome 7 were detected that had not been found by cytogenetic analysis. Additional information on aberrations of chromosomes I and 7 in mesothelioma was obtained by interphase cytogenetic analysis on paraffin-embedded sections of I3 cytogenetically characterized tumor samples. More heterogeneity in chromosome copy numbers was observed and new clones were detected in comparison to cytogenetic analysis. The modal and mean chromosome numbers determined by cytogenetic analysis were correlated with flow cytometric findings in 3 I mesotheliomas. A statistically significant positive correlation was found between the DNA indices and both modal and mean chromosome numbers ( p = 0.04 and 0.006 respectively). As in other tumors, the cytogenetic aberrations are numerous and thus of no direct diagnostic value in mesothelioma. However, some of the chromosomal findings correlate with prognosis. The copy number of chromosome 7 p-arms was found to be inversely correlated with clinical survival ( p = 0.02). Moreover, the patients with hyperdiploid mean chromosome number have shorter survival compared to patients with normal karyotype or hypodiploid chromosome number ( p = 0.0007). Supernumerary copies of chromosomes may lead to elevated expression of genes critical in the pathogenesis of mesothelioma. The lung content of asbestos fibers was determined from 16 patients. Losses of chromosomes I and 4, and structural aberrations involving lpll-p22 were found to correlate with high amounts of asbestos fibers in the lung tissue ( p = 0.0001, 0.003, and 0.009 respectively). Cells with these aberrations may have a growth advantage in the microenvironment produced by the inflammatory reaction caused by the fibers. Though the patient material was limited even in this study, it has formed a basis to molecular approaches on the pathogenesis of human malignant mesothelioma. This interesting malignancy deserves further studies that are aimed at the elucidation of human carcinogenesis in general and asbestos-induced malignancy in particular. April 1992

Tumour cell rheology-Experimental studies in vivo and in vitro on factors influencing tumour cell lodgement and survival in microvessels

U. NANNMARK Department of Anatomy, University of Goteborg, P.O. Box 33031, S-400 33 Goteborg, Sweden lntravasated tumour cells (TCs) are subjected to a host of factors that may influence their microvascular arrest and survival. Since undeformed TCs are spherical and often large, marked shape changes occur when TCs enter capillaries. It has, therefore, been proposed that cell mechanics plays a role in T C arrest and, hence, location of secondary tumour growth. It has also been hypothesized that large deformations forced upon the TCs may be one explanation to the fact that most TCs die rapidly after entry into the blood circulation. In the present investigation, the rheological properites of different TCs were studied in relation to cell morphology, deformability, arrest and survival in microvessels. The morphology of rat fibrosarcoma cells from culture (CTCs) and solid tumour (STCs) was analysed by TEM, their deformability measured in 6.5 p m pipettes and their hepatic lodgements studied with isotope technique. The same parameters were studied for cells from rat adenocarcinoma tumours ( ACCs). CTCs had a larger nucleus and mean cell diameter ( 14.9 pm), were stiffer and had a 4 times higher hepatic arrest than STCs (0 11.2 pm). The nucleus, cell diameter ( 13.9 pm) and rigidity of ACCs were slightly less than of the CTCs and the hepatic arrest was 2.5 times higher than for STCs. All cells had sufficient membrane ‘excess’ (membrane wrinkles and villi) for constant volume deformation into the pipettes without true expansion of the membrane. Instead, it appeared that a large nucleus was the main restricting factor for deformation and unhindered circulation in microvessels. To study the influence on T C viability of different entrance flow velocities, as might occur in different microvascular beds, isotope labelled TCs were perfused at 0.25 o r 2.5 mm/s through micropore filters with 5 prn pore size. There was no statistically significant difference in cell death between the high and low flow rates. Neither were there any indications in SEM analyses of TCs in the fiter pores that a high rate of initial cell deformation into microchannels should be more lethal t o TCs than a low deformation rate. Fluorescence vital microscopy of the mouse cremaster muscle microcirculation was used for studies of T C ( M C G I A A tumour) deformation and survival. To mice pre-treated with ethidium bromide (EB), which gives red fluorescence to damaged cells, TCs stained with acridine orange (AO) were infused i.a. Many of the TCs reaching the muscle were shunted to post-capillary venules through wide capillaries, while TCs entering narrower capillaries were usually trapped and extensively deformed (mean lengths 53.5 pm), thus requiring large apparent increases in surface area. lntracapillary T C death was recorded by changes from green( AO) to red- (EB) fluorescence. Calculations of deformed TCs (in vivo) and undeformed TCs (TEM), showed that, despite the large deformations occurring in vivo, all membrane ‘excess’ of the TCs trapped in the capillaries had not, in average. been consumed. However, these are average figures and some of the TCs that died (81”h) within 30min after arrest may still have died due t o membrane ‘overstretching’. i.e. from mechanical trauma. An isolated rat heart model (cell-free perfusion) was used t o analyse if muscle contractions would increase the death of circulating TCs. TCs were infused into the coronary circulation of beating and non-beating (calcium-depleted) hearts. Viable TCs in the effluent were counted and the ultrastructure and location of TCs trapped in the myocardium were studied. Twice as many

ABSTRACTS OF THESES

869

viable TCs were retrieved from non-beating ( 14%)) than from beating (7%) hearts. TEM showed that 82% of TCs trapped in beating hearts vs. 35% in non-beating hearts had signs of plasmalemmal fragmentation and disorganisation of the nucleus and cytoplasm with formation of intracellular vacuoles. Such a generalized cellular damage seems to be better explained by a biochemical insult to the cells than by a biophysical, i.e. mechanical, trauma. Hypothetically, the TCs were killed by toxic oxygen species generated by the endothelium. Production of these species is known to be calcium dependent, which might explain why T C killing was least efficient in calcium depleted, non-beating hearts. In conclusion, the mechanics of TCs influences their arrest in microvessels. TCs in culture and from solid tumour have different rheological properties, with cultured cells being larger and stiffer. Therefore, to best reflect the fate of spontaneously released TCs. TCs from solid tumour should be used in experimental lodgement studies. The rate of T C deformation does not influence T C viability. TC deformations in vivo may be extensive and possibly kill TCs by membrane rupture, but this mechanism is probably of limited importance. More TCs are killed in the microcirculation of beating hearts (calcium present) than in non-beating hearts (calcium depletion). This is proposed to be due to a biochemical insult (e.g. oxygen radicals from the endothelium) rather than to mechanical trauma.

13% corrected 5-year survival. Patients who underwent intestinal surgery as a part of initial surgical debulking had a very poor survival, even compared with the group of patients with greater than 3 cm residual tumors left after initial surgery ( 4 vs 13%)).The secondary laparotomy gave prognostic information only in stages I and 11. Twenty-six percent (18/68) of the patients who had macroscopic tumor left at the secondary surgery could be rendered tumor free at the second laparotomy. It was not possible to state that this was caused by the debulking per se. Geometrical measurements on the tumor nuclei o n the diagnostic tissue sections generated powerful prognostic factors for survival after secondary laparotomy in 65 patients with advanced ovarian cancer. The existence of very large nuclei seemed to characterize patients with a bad prognosis. The half-life of the tumor marker CA 125 in serum during induction chemotherapy gave equally good prognostic information regarding survival after secondary laparotomy as registering the response to therapy at the secondary laparotomy. Monitoring ovarian cancer patients with monthly serum CA 125 determinations during follow-up was a reliable method to diagnose a recurrence with very few (0.9%) false positive values.

April 1992

In situ carcinoma of the breast-Aspects on natural history and treatment with special reference to subcutaneous mastectomy A. RINCBERCHACBERC Dept of Plastic and Reconstructive Surgery, Malmo General Hospital, S-214 01 Malmo, Sweden

Ovarian cancer-Treatment results, prognostic factors, and tumor marker surveillance T. HOGBERG Departments of Gynecologic Oncology, Clinical Chemistry and Oncology, Linkoping University, S-581 85 Linkoping, Sweden

The total population-based material of 426 ovarian malignancies in the Southeast Health Care Region of Sweden during 1984-1987 was surveyed. After comparing our material with other epidemiological studies, it seems that with a program of cytoreductive surgery followed by a cisplatinum chemotherapy combination in the metastasizing cases the overall survival figures have improved in ovarian cancer. A relative overall 5-year survival of 43% was recorded. Patients aged under 45 had a 71% relative 5-year survival, and patients with FIGO stage I tumors 84'K Age and stage were independent prognostic variables for survival, while histology (epithelial vs. non-epithelial tumors) did not add prognostic information. Three hundred and eighty-four patients with ovarian carcinomas were analyzed separately. An overall relative 5-year survival of 40'%1was recorded. There was a pronounced policy to adhere to the treatment protocols, but still only half of the patients were considered fit for protocol treatment. The overall corrected 5-year survival for patients prescribed protocol treatment was 49% compared to 33% for those treated otherwise. Easily recorded clinical variables were used to create prognostic indices which divided the patients into risk groups. It was, however. shown that the high-risk patients mainly were those who actually did not get protocol treatment and that the prognostic indices were unsensitive among the high-risk patients who were prescribed protocol treatment. The surgical staging was found to be carelessly performed in the early stages of disease; only 27/148 (18'%~)of patients allotted to stage I or 11 underwent an adequate surgical staging procedure. The corrected 5-year survival for patients with FIGO stage 111-IV tumors was 35% if optimal primary cytoreductive surgery was performed. Patients with residual tumors greater than I cm had

April 1992

In situ carcinoma of the breast (CIS) was previously an infrequent finding. Two main forms have been described: ductal (DCIS) and lobular (LCIS) carcinoma in situ. Treatment was mastectomy (ME). With the introduction of mammography the proportion of CIS increased from a few to 10-20%, of all diagnosed breast carcinomas (BC). As a result of the introduction of breast conserving operations (BCO) as treatment for invasive BC. M E was questioned as treatment for CIS. The present study was undertaken to improve our understanding of the clinical significance of CIS and to define indications for subcutaneous mastectomy (SCM). It was found that after a biopsy diagnosis of CIS subsequent foci of malignant lesions, mainly of identical microscopical type, were found in the following SCM in a large number of patients. Extensive histological examination was performed. Similarly, malignant lesions were found in a high frequency in contralateral SCM specimens. Fewer bilateral lesions were seen if the primary biopsy diagnosis had been DCIS ( 17%) than if it had been LClS (63%). At autopsy of women with previous ipsilateral invasive BC contralateral malignant lesions were seen in 2 3 % ~The proportion of LClS was lower than in the clinical series, which may be attributed to age ditferences. In a demographic study in the city of Malmo the incidence of CIS seemed to be related to age. with the highest incidence in the pre- and perimenopausal years, and also to the breast diagnostic activities. Bilateral SCM was standard treatment for CIS in Malmo since 1979. It was shown that SCM was associated with substantial mental morbidity. Complications were common, with 1/4 of the patients needing a second operative procedure. At follow-up half of the patients had remaining sequelae (unsightly scars. improper implant position, hardening of the implant). Even if results have improved with time. SCM should only be performed on strict indications. It may be an alternative to ME for large DCIS lesions. In the present study the frequency of subsequent ipsilateral invasive BC and BC mortality after DCIS was in agreement with current data in the literature. lpsilateral invasive disease develops in about

870

ABSTRACTS OF THESES

5- 10%)after BCO with survival rates of 95-loo%, equal to those after ME. These results suggest that only a part of all DCIS lesions will develop to invasive carcinoma. The results support a more conservative approach than bilateral SCM. Controlled trials to study the outcome after BCO are important. Present results also support the view that LCIS is a marker of increased risk to develop invasive carcinoma and that follow-up after BCO may be appropriate. May 1992

Flow cytometric DNA content and S-phase fraction in endometrial adenocarcinoma- A descriptive a n d prognostic study. G. WAGENIUS Department of Oncology, Uppsala University, Akademiska Sjukhuset, S-751 85 Uppsala, Sweden Flow cytometric measurements of DNA content and S-phase fraction were performed in a prospective study of totally 406 women with invasive endometrial adenocarcinoma in stages I-IV. Seventy-six percent of the endometrial adenocarcinomas were peri-diploid ( 1.8cP2.2c) and 24 percent were aneuploid. The S-phase fraction varied between 3.1% to 40.8'Yo with a mean value of 12.0%~Aneuploidy and high S-phase fractions were significantly correlated to advanced clinical stage and poor histopathologic differentiation and aneuploidy was associated with postmenopausal status. Residual myometrial invasion after intracavitary treatment was more frequently found in aneuploid tumors. In a univariate analysis, age, uterine cavity depth, clinical stage, histopathologic grade and myometrial invasion, as well as aneuploidy and high S-phase fractions, were correlated to prognosis. In the multivariate analyses, histopathologic grade, clinical stage, myometrial tumor invasion, DNA content and S-phase fraction were independently correlated to prognosis, but provided that the S-phase fraction WAS measurable, DNA content and histopathologic grade did not give any additive prognostic information. Apart from clinical stages 111 and IV the S-phase fraction was the strongest predictor of survival. In the clinically interesting stages I and 11, only S-phase fraction and myometrial tumor invasion were statistically independent prognostic factors. S-phase fraction was found to be a time-dependent prognostic Factor with most prognostic importance during the first two-three years after diagnosis. Using multivariate analyses, six different prognostic indices were created for different pre- and post-treatment situations. With one of the indices the relative hazard for a given S-phase fraction and myometrial tumor invasion could be calculated, and with five of these indices a small high-risk group (those 10"h with the lowest index values) with a 3-year survival of 44-62% and a large low-risk group (those 90% of the patients with the highest values) with a 3-year survival of 88-96'Xr could be identified. The present results show that both ploidy level and S-phase fraction add prognostic information to established clinical parameters and that it is possible to identify relatively well defined high- and low-risk groups. May 1992 Mineral fibers, cigarette smoke, and oxidative DNA damage- An experimental study P. LEANDERSON Departments of Occupational Medicine and Clinical Chemistry Linkoping University, Faculty of Health Sciences, S-58 I 85 Linkoping, Sweden

This study concerns the ability of mineral fibers and cigarette smoke to generate reactive oxygen metabolites and to cause damage to DNA. The generation of hydroxyl radicals (OH') was demonstrated by using the DNA base, deoxyguanosine ( d G ) as a trapping agent for OH' and determining its hydroxylation to 8-hydroxydeoxyguanosine (80HdG). and oxidative D N A damage was assessed as formation of 80HdG in isolated DNA or in DNA of human lung cells in culture. Both asbestos (chrysotile) and manmade mineral fibers ( M M M F ) were found to generate OH' and damage D N A in aqueous buffer solutions, and more OH' was generated by the iron-containing rock and slag wools than by glass wool and ceramic fibers. The OH' generation by rock wool was decreased by treatment with heat, oxygen or desferrioxamine. indicating that chemical characteristics including iron o n the fiber surface were important determinants of the OH' formation. A variety of natural mineral fibers (asbestos, erionite, and wollastonite) and M M M F (rock wool, glass wool, and ceramic fibers) were found to stimulate polymorphonuclear leukocytes (PMNL) to generate OH' in the presence of exogenously added iron, and amosite, crocidolite, antophyllite, erionite, and wollastonite caused OH' formation also in the absence of exogeneously added iron. Cigarette smoke potentiated the damaging action of rock wool on isolated DNA, suggesting that iron-containing fibers might catalyze OH' formation from hydrogen peroxide generated in the smoke. Cigarette smoke-induced damage to isolated DNA was inhibited by tyrosinase and catalase, indicating that polyphenols in the smoke, e.g. hydroquinone and catechol, were important for the hydrogen peroxide generation. Cigarette smoke WAS also found t o cause oxidative DNA damage and DNA strand break formation in cultured human lung cells by mechanisms involving OH' attack on the DNA molecule and endonuclease activation. Moreover, cigarette tar was demonstrated to promote PMNL-mediated DNA strand-break formation in human lung cells, and tar loaded with iron was more damaging than regular tar. These findings indicate that mineral fibers, by producing OH' themselves and by stimulating PMNL to generate OH' in the presence of iron, may cause oxidative DNA damage under experimental conditions. They also indicate that cigarette smoke may cause DNA base hydroxylation and DNA strand-break formation in human lung cells via mechanisms involving OH'. and that iron is important for the OH'-formation. Altogether, the findings point to the possibility that mineral fibers may promote OH' generation by cigarette smoke and inflammatory cells and so increase the risk of DNA damage in human lung cells. May 1992

Cellular effects of acrolein and fecapentaene-12 associated with multistage carcinogenesis C. C. UNGER Department of Toxicology, Karolinska Institutet, S-104 01 Stockholm, Sweden The present study has primarily utilized human cell cultures t o study cytopathogenic effects of two reactive agents, i.e. the unsaturated aldehyde acrolein and fecapentaene-12. The latter agent is a mutagen produced by colon bacteria and believed to decompose to unsaturated aldehydes and oxyradicals. Acrolein markedly decreased colony survival and membrane integrity of bronchial epithelial cells and fibroblasts, and moreover, depleted cellular glutahione and protein thiols. Furthermore, acrolein was genotoxic and caused both DNA single strand breaks and DNA protein crosslinks. Acrolein also induced terminal squamous

ABSTRACTS OF THESES

differentiation of the epithelial cells measured as decreased growth and increased formation of crosslinked envelopes. Because acrolein is more cytotoxic and mutagenic to skin fibroblasts obtained from individuals with xeroderma pigmentosum than to cells from normal individuals, the role of cellular thiols and the formation and possible repair of DNA single strand breaks were studied in these cell types. Acrolein decreased the levels of reduced glutathione and free protein thiols by similar kinetics in both cell types. However. acrolein caused higher loss of membrane integrity and a higher extent of DNA single strand breaks in the XP cells that in the normal cells, possibly related to the fact that normal cells contained comparatively higher amounts of free thiols. Exposure and subsequent removal of acrolein resulted in continued formation of DNA single strand breaks in the normal cells and these lesions accumulated to a similar extent as in the presence of agents which efficiently inhibit DNA repair synthesis. Fecapentaene-I2 was also cytotoxic as measured by loss of colony survival and membrane integrity of human skin fibroblasts and decreased the content of cellular glutathione. Several types of cellular DNA damage were induced by this agent including DNA single strand breaks, DNA protein crosslinks and DNA interstrand crosslinks. In mixtures with either glutathione or plasmid DNA, fecapentaene-12 reacted with glutathione indicating both alkylation and oxidative reactions, or caused the similar DNA lesions detected in cells. Taken together, these studies indicate that acrolein and fecapentaene cause effects in vitro that relate to carcinogenesis as a multistep process. Moreover, a relatively similar spectrum of cellular effects from both agents indicate that aldehyde formation may primarily underlie the cytotoxic and genotoxic actions of fecapentaene-12. May 1992

Oral contraceptives and breast cancer J. RANSTAM Department of Community Health Sciences, Malmo General Hospital, 5-21401 Malmo, Sweden

Oral contraceptives (OCs) have in this study been found to increase the risk of breast cancer. The exposure-response relationship appeared to be modified by age at exposure. O C usage before 20 years of age seemed to be related to higher risks than usage between 20 and 25, and at higher ages the effects appeared to be negligible. A high proliferation rate. a large proportion DNA-aneuploidy, and amplification of the proto-oncogene HER-2/neu appeared to be related to OCs when used during the adolescence but not to OC usage at higher ages. The prognosis appeared to be worse among patients with a history of OC usage during the adolescence. The incidence of breast cancer has been found to be increasing in Sweden during the late seventies and the eighties. Presently, usage of OCs is widespread and to a high degree initiated during the adolescence. All relationships with OCs in this study relate, however, to brands that were used during the sixties and early seventies only. Modern brands of OCs have a substantially reduced dosage of steroids. Whether the new brands of low-dose pills are safe with respect to breast cancer is presently unknown. Moreover, it cannot yet be studied epidemiologically because such studies will not be reliable until those exposed in their adolescence, during the late seventies and eighties, reach 40-50 years of age. It is also unknown if the presented effects on breast cancer risk and prognosis apply to women below 50 years of age only, or if they persist at higher ages. It remains, thus, to be assessed if the

87 1

increased breast cancer risks, related with past and present OC usage, will have major implications on the public health. May 1992

Bleeding complications and coagulopathy in acute leukemia with special emphasis on proteolytic degradation by elastase E. TORNEBOHM Division of Clinical Hematology and Oncology, Department of Medicine, Karolinska Institute, Huddinge University Hospital, S-141 86 Huddinge, Sweden

Bleeding complications in the course of acute leukemia are still a major clinical problem despite a liberal platelet transfusion policy. However, bleeding diathesis can occur at any platelet level. Aims: This thesis had three aims. First, to study the incidence of hemorrhage in a large patient material with acute leukemia. Secondly, to more specifically study the coagulopathy to obtain a better understanding of the different causes of major bleedings. Thirdly, to investigate the potential role of the proteolytic enzyme elastase in bleeding complications. Patients and rneth0d.y: Hemorrhagic symptoms on admission, during induction chemotherapy and as a cause of death were investigated in 259 initially chemotherapy-treated patients with primary and secondary acute leukemia. Of these 18 were bone marrow transplant patients; an additional 179 patients. referred for transplant, were also studied. Coagulation and fibrinolytic variables were analysed in 72 patients at diagnosis. With a method, developed to measure intracellular elastase in leukemic cells, 60 patients with de novo acute leukemia were investigated. Results: Initial bleeding symptoms were present in 38% of the patients, most frequently seen in promyelocytic leukemia (M3). Hemorrhage on admission had n o prognostic significance. The early hemorhagic mortality was 10% and in the majority due to intracranial hematomas and gastrointestinal bleedings. N o early lethalhemorrhage was seen in lymphoblastic leukemia (ALL). Besides thrombocytopenia the initial leukocyte count was found to be an independent risk factor for bleeding complication during the first month. The risk of developing a hemorrhagic complication increased significantly at leukocyte levels > 2 0 x 10y/l. The total hemorrhagic mortality was 23% and no decline in the incidence was observed during this 20-year period. Lethal bleeding complications after bone marrow transplantation were seen in 19%. The majority of these were patients treated with allogenic grafts and in hematological reconstitution. Graft-versus-host disease and septicemia were common factors contributing to death. Twenty-seven of 72 patients had bleeding complications. In all 72 patients coagulation disturbances were frequent and certain characteristics were observed for different subtypes of acute leukemia. A fibrinolytic state with low concentrations of fibrinogen, antiplasmin, high fibrin (ogen) degradation products and normal antithrombin levels was especially evident in the M3 patients. Low concentrations of antithrombin were seen in the patients with myeloblastic and monocytic leukemia, the latter also frequently had low antiplasmin levels. Some of the ALL patients had extremely high ratios of Von Willebrand Factor antigen/Factor Vlll (vWF:Ag/F VIII). The patients with bleeding complications had significantly lower platelet counts, prothrombin complex, fibrinogen and protein C than those without. The ratio vWF:Ag/F VIII was a platelet-independent risk factor for hemorrhage in the group of patients with platelet levels < 25 x 10y/l ( p = 0.02). Above this level antiplasmin. fibrinogen and protein C were found t o be significant risk factors ( p < 0.05). Detectable intracellular elastase was found in 92%)of the myeloic leukemia patients. while

872

ABSTRACTS OF THESES

the ALL group lacked this activity, the M3 patients had very high levels. The total circulating leukemic cell elastase (TCLE) activity in the tumor load was estimated and 12 patients from different myeloic subtypes (M3 included) showed very high levels. The ANL patients with hemorrhage had a more frequent incidence of elevated TCLE (70%) compared with the group without (36%. p < 0.05). There was a positive correlation between TCLE/L and plasma levels of elastase complexed with its inhibitor, q - a n titrypsin ( E - a , -AT). The patients with elevated levels of E-a,-AT had a significantly lower concentration of prothrombin complex, F VII, protein C and antiplasmin together with an elevated ratio vWF:Ag/F VIII ( p < 0.05). Conclusions: The tumour load mirrored by the leukemic cell count seems to be of importance in the development of hemorrhagic complications. Cytostatic treatment may result in a release of proteolytic enzymes from the leukemic cells. This can induce coagulation disturbances of clinical importance especially in promyelocytic leukemia, which had a several 1000-fold increase in intracellular elastase content compared with a normal neutrophil granulocyte. May 1992

Blood coagulation inhibitors in malignant disease with particular emphasis on tissue factor pathway inhibitor and its interaction with heparin A. K. LINDAHL Haematological Research Laboratory and Deparment of Surgery, Aker Hospital, University of Oslo, Oslo, Norway

In this study we have found that patients with solid malignant tumors have normal levels of coagulation inhibitors at time of diagnosis. With progression of disease, the plasma levels of antihrombin, heparin cofactor 11, protein C and protein S decrease, probably as a result of the combined effect of consumption and reduced synthesis by the liver. There is evidence of increased degree of coagulation activation even in patients with localized tumors. In patients with metastatic cancer, the degree of hypercoagulation is further elevated. We found negative correlations between the levels of the liver produced coagulation inhibitors and the degree of hypercoagulation. Both the inhibitor levels and the degree of hypercoagulation were correlated with the extent of the malignant process. The levels of tissue factor pathway inhibitor (TFPI) increase progressively with advance of the malignant disease. The increased levels of TFPI seen in patients with metastatic cancer is probably due to increased synthesis of TFPI. Which cells and what kind of mechanism could cause such increased synthesis is still unknown. TFPI is responsible for the indirect anticoagulant effect of heparin; the postheparin effect. In patients with metastatic cancer, more than half the anticoagulant effect of heparin may be mediated by TFPI. This is both due to the increased TFPI levels and the increased heparin-mediated TFPI release. The present study showed that patients with metastatic cancer not only have increased total levels of TFPI. They also have increased amounts of TFPI with particularly high anticoagulant effect. One might speculate that there would have been more thromboembolic disorders in cancer patients if the TFPI levels were normal. Exactly what role this additional anticoagulant TFPI effect plays in the interaction between hemostasis and malignancy is not known. May 1992

Late-effects after treatment for germ-cell cancer with cisplatin, vinblastine, and bleomycin S. W. HANSEN Department of Oncology B and Department of Clinical Physiology and Nuclear Medicine, The Finsen Institute, Rigshospitalet. DK-2100, Copenhagen, Denmark

During a 5-year period from 1979 to 1983 all patients in Denmark with metastatic non-seminomatous and extragonodal germ cell cancer were treated with 6 cycles of cisplatin, vinblastine. and bleomycin (PVB). Thirty-nine patients referred to the Finsen Institute accepted a follow-up examination of side-effects 3.5-9 years after chemotherapy. Renal toxicity consisted of an irreversible decrease in glomerular filtration rate ( G F R ) in 47% of the patients, while the decrease in G F R was fully reversible in 23‘%1. Significant pulmonary toxicity was observed in smokers and consisted of an irreversible decrease in carbon monoxide diffusion capacitiy to median 72% of the predicted value. Neurotoxicity was the most pronounced long-term side-effect. Nearly all patients had a peripheral sensory neuropathy probably caused by axonal degeneration. A central conduction defect was observed in 88% of the patients by measuring auditory brain-stem potentials. Irreversible high-frequency hearing loss was induced in 39% of the patients. Parasympathetic nerve dysfunction was found in 36% of the examined patients, including 2 patients with impotence. Half of the patients revealed Raynaud’s phenomenon (RP), and the mechanism underlying this side-effect was found to be hyperreactivity of the central sympathetic nervous system. Vascular toxicity was found only in terminal arterioles and was not responsible for RP. PVB treatment caused low sperm counts and a subclinical Leydig cell dysfunction in the majority of patients. Azoospermia was observed in 27% of the patients. Six patients had hypertension and this was not related to renal impairment. Objective measurements revealed several long-term complications, which for the most part proved to be without clinical signficance at the time of follow-up. May 1992

Occupational health risks in high voltage cable workers exposed to oil mist and electromagnetic fields K. SKYBERG Medical Department, Alcatel STK AS and National Institute of Occupational Health, Oslo, Norway

This research describes an increased risk of pulmonary fibrosis and lung cancer among workers exposed to mineral cable oils by inhalation. Health hazards from present oil exposure levels have been discussed. A possible adverse health effect from exposure t o electromagnetic fields and electric spark discharges should be further investigated. Although there are certain limitations with respect to the exposure estimates and the size of the exposed populations, it seems reasonable to draw the following conclusions: There has been a high risk of developing pulmonary fibrosis among workers employed in oil filled cable production at the old Oslo plant. Among workers employed during 1963- 1980 and exposed for 10 years or more the risk was 50%. The bellows function of the lungs was unaffected, but especially among workers employed for 10 years or more, the C O transfer factor was decreased, and during maximum exercise, arterial oxygenation was reduced, although ventilation was increased. Impregnation workers reported respiratory symptoms more frequently than controls. Inhalation exposure to mineral cable oils is the most plausible cause of the fibrosis. but asbestos exposure and smoking may have contributed.

ABSTRACTS OF THESES

Modern production facilities and hygenic measures have reduced the total oil concentration in oil filled cable impregnation from an estimated 50-100mg/m' to a measured level of 1030 mg/m'. and in this situation no cases of oil related pulmonary fibrosis have occurred after 14 years of exposure. Considering that the oil mist levels (predominantly high viscosity oils) were similar in the impregnation and lead sheathing departments at the old Oslo plant, and only one case of fibrosis was observed among 11 sheathing workers, it seems that exposure to oil vapour at a level of 50- 100 mg/m3 (predominantly low viscosity oils) is responsible for the fibrogenic effect on the lungs. Oil filled cable production workers at the Oslo plant and installation workers were at higher risk of dying from ischaemic heart disease. The risk increase was statistically uncertain, and may partly be explained by smoking. Although there was a lower risk of getting cancer at any site than expected, the risk of developing lung cancer was two to three times the expected risk. Exposure to mineral oils is considered to be an important contributing factor in the development of lung cancer. A moderate asbestos exposure, especially in the sheathing department, may have contributed to the increased risk. In the departments where the risk was increased, total oil concentration levels were 10-100mg/m3 and oil mist levels 0.5-4mg/m3. Due to the relatively small size of the cohort, the ability of the study to detect moderately increased risk of cancer in other organs was limited. The study does not provide information about the lung cancer risk after 1964, taking into account a latency period of at least 20 years from first employment in oil exposed work. An extended follow-up period is important, since the production of oil filled cables at the Oslo plant was at the highest during 1960-1979. A cytogenetic study of oil exposed cable workers in 1984 showed no oil related genotoxic effects, but this does not rule out a possible cancer risk from present exposures. However, both skin and inhalation exposure has been reduced, total oil concentration is typically 5-10 mg/m3. and this is expected to reduce the risk of lung cancer. The experimental studies have shown that the inhalation toxicity of different cable insulation fluids are quite different. Mineral oils still in use have been related to an increased cancer risk. The experimental inhalation study has indicated that one mineral oil and some synthetic alternatives may be toxic to the lungs. One synthetic cable fluid seems to be less biologically active than mineral oils. Inhalation studies of longer duration are desirable. before it is decided to replace the most commonly used mineral oil by a synthetic alternative. The cytogenetic studies indicate that exposure t o electromagnetic fields and spark discharges is related to chromosome damage in peripheral lymphocytes. This finding is supported by one other epidemiological study and several in vitro experiments. The data indicated an interaction between exposure and smoking. There is some reason for concern, since other research implies a relation between genotoxic effects in the peripheral lymphocytes and future cancer risk (Brngger et al. 1990). Unnecessary exposure to electromagnetic fields and spark discharges should therefore be avoided. Smoking cessation should be promoted. June 1992

Trace elements in surgical and oncological patients- A prospective, clinical study on subjects undergoing variously straining therapies H. ANTILA Departments of Anaesthesiology and Clinical Chemistry. University of Turku. Turku. Finland

873

The effects of surgical and oncological therapy on the common parameters of trace element and nutritional status were studied in 64 patients to detect development of any disturbances in trace element status. Reference values for mononuclear and polymorphonuclear leukocyte zinc content as well as serum trace element concentrations were determined with the particle-induced x-ray emission method used in the study in a group of 199 healthy subjects. The patient groups of the study were selected to represent treatments causing various degrees of tissue damage induced by operation, irradiation or cytotoxic therapies. They consisted of 10 patients undergoing coronary bypass operation, 17 patients with postoperative intermaxillary fixation, 24 breast cancer patients receiving postoperative radiotherapy and I3 bone marrow transplant recipients on TPN. Eight patients with intermaxillary fixation received additional supplementation with a commercial enteral formula to meet their basal trace element requirements. All bone marrow transplant recipients received total parenteral nutrition with trace element substitution as recommended by the American Medical Association. The following conclusions were drawn: -

-

-

-

The reference values for mononuclear and polymorphonuclear leukocyte zinc content as well as serum trace elements were in line with the results of other recent studies. The copper, selenium, iron and bromine of mononuclear and polymorphonuclear leukocytes could not be reliably analyzed by the current methods used in this study. Mononuclear and polymorphonuclear leukocyte zinc content may give additional information about zinc status. However, the separation of cells is time-consuming and costly. Consequently, their use in clinical practice has to be directed to patients, in whom serum zinc levels are of limited value, e.g. patients in the immediate postoperative period or those with infection. Furthermore, they can be used when subclinical zinc deficiency is suspected but serum zinc concentrations are within the normal range. N o clinical signs of trace element deficiency syndromes were seen in any of the patient groups. However, after major surgery, serum zinc and transferrin bound iron were even at two months after operation significantly lower than the initial values. This may reflect subclinical deficiency, since the actute reaction and immediate postoperative period were already over. The high serum selenium levels in the two patient groups from 1987- 1989 reflect the increased selenium intake of the Finnish population due to enrichment of fertilisers with sodium selenate. The trace element supplementation in current use is sufficient to prevent clinical signs of deficiency and to maintain serum levels within reference values in bone marrow transplant recipients receiving total parenteral nutrition. The significance of the low initial serum zinc levels seen in these patients should be analysed in further studies. Oral supplementation with a commercial formula helps to preserve the weight of patients with limited surgical trauma and impaired oral intake. The formula used in the study did not contain sufficient amounts of trace elements to cause any difference in trace element parameters between the groups.

June 1992

Endocrine treatment of prostatic adenocarcinoma- An experimental study in the rat. M. LANDSTROM Departments of Physiology, Pathology and Urology & Andrology. University of Umel, S-901 87 Umel, Sweden

874

ABSTRACTS OF THESES

The present study was designed to evaluate and compare effects of different endocrine treatments in two rat prostatic tumor models (Dunning R3327 and AT-I). The aim was to investigate if treatments with estrogens or gestagens affect tumor growth and morphology, by mechamisms not related to the suppression of androgens. Treatment of castrated rats bearing prostatic adenocarcinomas with 50 p g estradioLl7P (E2) S.C. daily inhibited tumor growth more effectively than castration alone, and the number of epithelial cells was reduced. Castration alone decreased the tumor epithelial cell size, while the number of cells was the same as in tumors from intact animals. Anti-estrogen, tamoxifen, did not block the inhibitory effect of E2 on tumor growth rate in castrated rats. suggesting that the effect of E2 is not mediated by interaction with estrogen receptors on the tumor. In a doseeresponse study on the effects of E2 in castrated and testosterone-supplemented rats, a dose of 1Opg S.C. daily or higher was needed to inhibit growth. In intact (non-castrated) animals, a dose of 1 p g E2 S.C. daily was sufficient to inhibit the tumor growth rate. These findings indicate that the suggested direct effect of estrogen on prostatic tumors is present at a 10-fold higher dose level than the anti-gonadotrophic effect of estrogen. In order to study the effect of E2 on the tumor growth rate in the hormone-insensitive state, two tumor models were used: the Dunning R3327 relapsed after castration and the androgen-independent, anaplastic Dunning AT-I. In both models E2 treatment suppressed the tumor growth rate and decreased the tumor mitotic index. In the relapsed Dunning R3327 model, E2 partly prevented the dedifferentiation and increased the number of apoptotic cells, suggesting an increased cell death rate. The present work also demonstrated that estrogens had stimulatory effects on the stromal compartment of prostatic tumors. In an attempt to clarify the stimulatory effects of E2, the occurrence of different cytoskeletal and extracellular matrix proteins was studied using immunohistochemistry. The amounts of collagens I and 111 and fibronectin were considerably elevated in the tumor stroma after E2 treatment. Cytokeratin 14, a marker for basal epithelial cells, was reduced by E2 treatment of castrated rats. Two gestagens, medroxyprogesterone acetate (MPA) and 6methylene-4-pregene-3,20-dione(6MP), were used to treat experimental prostatic adenocarcinoma, and both substances had growth inhibitory effects on the tumor. Combined treatment with E2 and MPA caused more inhibition than E2 treatment alone. These results support the idea that gestagens have direct inhibitory effects on prostatic adenocarcinoma. It was also shown that the decrease in tumor blood flow after castration was conteracted by treatment of the animals with 6MP. The results in the present thesis show that estrogens and progestagens have tumor-inhibitory effects on prostatic adenocarcinoma which are not related to suppression of androgens. June 1992 Cell proliferation rate in prostatic carcinoma: prognostic value and relation to epidermal growth factor receptor and pJ3 expression

T. VISAKORPI Department of Clinical Chemistry, Tampere University Hospital and Department of Biomedical Sciences, University of Tampere. Tampere, Finland This present study investigated the prognostic value of cell proliferation rate in prostatic carcinoma with DNA flow cytometry and PCNA immunohistochemistry. The association of p53. EGFR and ERBB2 expression with cell proliferation and prognosis was evaluated.

The use of D N A histogram background subtraction t o correct the influence of sliced nuclei improved the prognostic value of the cell proliferation rate defined as flow cytometric SPF. Good reproducibility of the cell cycle distribution analysis and improved prognostic value of SPF warranted the use of background subtraction in flow cytometric measurement of cell proliferation. Cell proliferation rate determined by flow cytometry was an independent prognostic factor in multivariate analysis of MO-stage patients. A very high cell proliferation rate also predicted a poor response to endocrine therapy for patients who received solely endocrine therapy. DNA aneuploidy, which predicted a poor prognosis in univariate analysis, failed to yield additional prognostic value when cell proliferation rate, histological grade, and T-stage were taken into account. Immunohistochemical analysis of PCNA using either PClO or 19A2 antibody was not associated with the cell proliferation rate determined by FCM. PCNA analysis with 19A2 antibody was marginally associated with prognosis, while analysis with PClO antibody revealed no prognostic value at all. The prevalence of p53 accumulation was low, indicating that the impact of p53 mutation on the pathogenesis of prostatic carcinoma is marginal. P53 accumulation was associated with poor tumor differentation and a high cell proliferation rate. It also predicted poor prognosis, although it was not an independent prognostic factor in multivariate analysis. A small subgroup of prostatic carcinomas lacking the EGFR expression was found. These tumors are well differentiated and proliferate slowly. The prognosis for patients with EGFR negative tumor was good, although EGFR expression was not an independent prognostic factor in multivariate analysis. Only a small group of prostatic carcinomas showed a very low level of ERBB2 immunostaining. This staining probably did not describe the expression of ERBB2 oncoprotein at all. It seems that ERBB2 overexpression is not important in the pathogenesis of prostatic carcinomas. The study indicates that flow cytometric analysis of the cell proliferation rate gives independent information on the prognosis for MO-stage prostatic carcinoma. The expressions of EGFR and mutated p53 are both involved in dysregulation of the cell proliferation rate of prostatic carcinomas, but are not as strong prognostic factors as determination of the cell proliferation rate itself. Flow cytometric cell proliferation ,measurements could help to detect the aggressive prostatic carcinomas and thus probably help in the selection of the most effective therapy for individual prostatic carcinoma patients. June 1992

Studies on the uptake of L-lmethyl-"Clmethionine in cancer with positron emission tomography

S. LESKINEN-KALLIO Department of Oncology and Radiotherapy, and Turku Medical Cyclotron/PET-Center. University of Turku. Turku, Finland

Uptake of ["Clmethionine in human cancer was studied by PET in order to elucidate the role of ["Clmethionine in detecting malignant tumors and assessing the proliferation activity of cancer. Four different analysis methods of the uptake of ["Clmethionine in cancer were studied. ["CIMethionine accumulated avidly in 57/61 (93%) malignant non-brain tumors (lymphoma, breast cancer. and head and neck cancer). The reason why 4 tumors were not visualized is their small size (diameter < 20 mm) and small proliferation rate. There was some accumulation of ["Clmethionine in an abscess, but no

ABSTRACTS OF THESES

accumulation in an inflammatory lesion or a benign breast tumor. The uptake of ["Clmethionine in low grade lymphomas was marked, while the uptake rate of F D G was low in 5/10 low or intermediate grade lymphomas. The uptake rate from the plasma to the lymphoma tumor was lower for F D G than for [ "Clmethionine. The uptake rate ( K , ) of ["Clmethionine from the plasma,to the tumor can be measured according to a graphical method of Patlak et al. However, the radioactivity accumulation in tumor adjusted for the injected dose and body surface area ( R D A ) correlates strongly with the uptake rate K,(r = 0.92, p < 0.0001). For clinical purposes, RDA and K, are comparable ways to analyze the accumulation of ["Clmethionine by PET, but RDA is simpler to derive, especially since blood samples are not needed and the dynamic scanning time is reduced from 40 min to 5 min. The radiation exposure to the personnel is also reduced when frequent blood samples need not to be taken. The capacity of the PET unit is improved. A high uptake K, of ["Clmethionine in the tumors of 22 patients tended to be associated with a large S-phase fraction ( r = 0.40, p = 0.07, n = 22). except for one case, where the uptake was clearly higher than predicted by the size of the S-phase fraction ( r = 0.62, p = 0.0003. n = 21, outlayer excluded). The correlation between the uptake rate of ["Clmethionine in DNA aneuploid tumors and the S-phase fraction was good ( r = 0.73, p = 0.005; for RDA 0.60. p = 0.03). The correlation between the uptake rate of ["Clmethionine from the plasma to the tumor ( K , ) and the S-phase fraction and between the accumulation adjusted for the injected dose and the body surface area (RDA) and S-phase fraction were of the same order of magnitude. The method could be valuable in certain clinical situations, e.g. for the assessment of the proliferation rate of deep-seated tumors and in cases where taking a surgical biopsy is not possible. June 1992

T cell-mediated immune recognition of human tumors regulated by MHC class I and ICAM-I molecules P. WANG Department of Tumor Biology, Karolinska Institute, S-104 01 Stockholm. Sweden The aim of this thesis was to study the expression of M H C class I and ICAM-I molecules on ex-vivo tumor cells and the function

875

of these molecules in the recognition of tumor cells by T lymphocytes in vitro. Tumor cells express relatively low level of M H C class I molecules. By isoelectric focusing analysis selective loss of certain class I allele products was detected in 23 of 50 carcinomas. Selective defects randomly occurred within the HLA, A, B and C locus. This selective downregulation of class I molecules may have functional consequences, because the M H C class I molecules present antigen epitopes to cytotoxic T cells. In one ovarian carcinoma the assembly of heavy and light chains was defective. One HLA A2 binding peptide and IFN; could correct this defect and induce the expression of class I molecules. The function of M H C class I molecules of immune response against tumor cells was investigated by generation and characteristics of CTLs in MLTCs. Clonal expansion of autologous tumor specific C D 8 + CTLs in 1/21 MLTCs was induced by repeated stimulation with autologous tumor cells. The lytic function of these CTLs was restricted by HLA B5. These results demonstrated that M H C class I molecules are essential, but not sufficient, for generation of immune response to tumors. By clonal analysis of MLTC responders we found that both CD4 and CD8 clones could express auto-tumor specific and class I restricted lytic function. In addition to earlier findings that the C D 4 T cells function as helper or suppressor, the class I restricted CD4 + CTLs reflected the heterogeneity of CD4 + T cells in response to class I positive and class I1 negative tumor cells. It has been described that TCR has low affinity interaction with MHC-antigen. Therefore, adhesion molecules are required for generation of cellular immune response. The induction of ICAM-I expression on tumor cells, which carried class I antigens, could elevate the auto-tumor recognition. The regulatory function of ICAM-I on the lymphocytes was also studied. The results of the existence of lytic functions in ICAM-1 positive lymphocytes and reduction of lytic function by pretreatment of lymphocytes with anti-ICAM-l mAb indicated that the ICAM-I o n lymphocytes could promote and regulate the functional activation of lymphocytes when it interacted with the LFA-1. These results demonstrated that M H C class I molecules, which are frequently altered in solid tumors, were important for generation of auto-tumor response in vitro. Expression of ICAM-I molecule on both tumor and lymphocytes could increase and/or regulate the immune response against tumor. October 1992