Keynote addresses

Radiation Oncology, Biology, Physics

110

Volume 15, Supplement 1, 1988

KEYNOTE ADDRESSES MONDAY,OCTOBER 10, 1988 MALIGNANT GLIOMA RADIATION ONCOLOGY

A NEMESIS

WHICH

REQUIRES

CLINICAL

AND

BASIC

INVESTIGATION

IN

Lawrence W. Davis, M.D. Montefiore Medical Center/Albert Einstein College of Medicine, Bronx, New York Malignant gliomas account for 40% of all central nervous system malignancies. These tumors are essentially a locali~ed neoplastic process which has defied most treatment . In spite of improved techniques, surgery is unlikely to increase survival further since true cancer operations cannot be performed. Radiation therapy has made a significant difference in outcome. Investigation in radiation oncology is essential for further improvement in the treatment of these tumors. The pattern of failure is local tumor recurrence, but the method to overcome this resistance to treatment is not clear. Radiation therapy techniques and inherent radio-resistance have been considered as possible reasons for failure. With newer imaging procedures, the extent of tumor can be more accurately defined allowing improved treatment planning. Identifying an effective treatment program is more difficult, though. Studies have documented the beneficial effect of radiation therapy, but the optimal total dose or fractionation schedule has not been determined. While some studies have reported improved survival using higher radiation doses, others have reported no benefit. More recently, studies of multiple daily fractionation schedules have been conducted using two or three daily fractions. Equally confusing results have been reported. Histologically, these tumors have necrotic areas and may be radioresistant due to hypOXic cells. Treatment methods designed to overcome the radioprotective effect of hypoxia have yielded disappointing results. The addition of hypOXic cell sensiti~ers has not produced the expected improvement in outcome. Studies using neutron radiation therapy report tumor control but not improved survival. Radiobiologic information is now available which may contribute to our understanding of the response of these tumors to radiation. Further laboratory and clinical investigation is required. Carefully designed clinical trials are needed to test new treatment concepts, and all radiation oncologists should be prepared to participate in such clinical studies.

TUESDAY

I

OCTOBER 11, 1988

THE DEVELOfMEN'l' OP COMPtrrBR ALGORITHMS FOR RADIATION TREATMEN'l' PLANNING.

J.R. Cunningham, Ph.D. Chief Clinical Physicist, Ontario Cancer Institute, Toronto Professor, Department Medical Biophysics, University of Toronto The procedures of radiation therapy include a number of steps. One of these is the prediction of the radiation dosage pattern that will result when a particular arrangement of radiation be~ or sources is chosen for a treatment. There are two reasons for carrying out the calculation of this pattern in detail. The chief one is to design the best treatment for the patient. The other is to record what has been done so that post-treatment analysis can be carried out. Both are important. Evidence has been accumulating which suggests that the tissue response V~rSU5 dose curve is relatively steep and that the dose delivered to tissues in a radiotherapy procedure should be known to an accuracy of about is\. This is really quite a demanding requirement and if it is to be even nearly achieved, all of the steps included in the whole process must be taken with care. The dosage calculation is only one of these steps but it is an essential one. The problem of calculating the dose, at a point in a patient, from either photon or electron beams is extremely complicated, sufficiently so that no exact mathematical solution to it has been found. The story of the development of methods to calculate doses for radiotherapy is therefore a history of the search for better and better apprOXimate methods. Because all such calculations produce apprOXimate values and because Some approximations are better than others it is important that the treatment planner should have some understanding of the methods in use in Order that their applicability to a given problem might be assessed.

Proceedings of the 30th Annual ASTRO Meeting

111

WEDNESDAY, OCTOBER 12, 1988 TUMOUR HYPOXIA; G.E.

FRIEND OR FOE?

Adams

Professor

~

Director.

MRC Radiobiology unit,

Chilton,

Didcot,

Oxon,

OXll ORO.

England.

The oxygen content of malignant tissue critically affects its metabolic status. As a consequence of this, oxygen deprivation can confer resistance not only to radiation but to a variety of drugs also, thereby adversely affecting response to therapy. Conversely, hypoxia can provide the basis for developing new and improved therapeutic strategies based on the metabolic activation of bioreductive drugs, some of which can also act as radiation sensitizers. Hypoxic radiation resistance in some tumours remains a topical area of research and new approaches to the problem of novel sensitising drugs will be described. The efficiency of bioreductive drugs both as cytotoxic agents and radiation sensitisers can be further enhanced by deliberately manipUlating the severity of tumour hypoxia. This can be achieved selectively by using agents that influence oxyhaemoglobin association and by various drugs that affect blood flow in tumours and the oxidative status of tumour cells in vivo. The results of experimental studies on tumours both of murine and human origin will be used to illustrate how the degree and severity of hypoxia can be manipulated to advantage. Potential applications exist for improving both diagnostic and therapeutic strategies. P-31 nmr spectroscopy is a versatile method for studying d~rectly and non-invasively the influence of these agents on the metabolic status of tumour tissue. Collaborative nmr stUdies using experimental solid tumours have shown that vasoactive drugs induce anaerobic metabolism. Substantial changes occur in the phosphorous-31 spectrum within minutes of drug administration and persist long after drug clearance. These observations are consistent with the finding that such agents can produce close to lOOt. radiobiological hypoxia.

THURSDAY, OCTOBER 13, 1988 LUNG CANCER CELL LINES, AUTOCRINE GROWTH FACTORS, CHROMOSOMAL DELETIONS, AND ONCOGENE ACTIVATION IN THE STUDY OF THE BIOLOGY AND PATHOGENESIS OF LUNG CANCER J. Minna, J. Battey, M.J. Birrer, J. Carmichael, F. Cuttitta, J. De Greve, A. Gazdar, E. Glatstein, J.W. Harbour, D.C. Ihde, B. Johnson, F. Kaye, G. Krystal, L. Lai, 1. Linnoila, J. Mitchell, J. Mulshine, M. Nau, H. Oie, E. Russel, E. Sausville, J. Schutte, T. Takahashi, F. Thomas NCI-Navy Medical Oncology Branch, National CancerInstitute, NavalHospital, & Uniformed Services University of the HealthSciences, Bethesda MD 20814 We have established a large panel of lung cancer cell lines of all histologic types from biopsies of tumor specimens from patients entering ontoour clinicaltreatment protocols. Theseprotocols for bothsmallcell (SCLC) and non-small cell lungcancer(non-SCLC) prospectively investigate the correlation between in vitro chemosensitivity responses of theselung cancerlines with the response of the patient's tumorto administered chemotherapy. We havefound: 1.) that we can establish lung cancercell linesfrom 40- 50% of the adequate tumorbiopsy specimens; 2.) that untreated SCLCis more sensitive in vitro than non-SCLC, or SCLClinesfrom patients who haverelapsed on chemotherapy; 3.) that thereis considerable heterogeneity of chemotherapy sensitivity between the lung cancerlinesof any histologic type. The resultsof clinicalcorrelations are early but suggestsignificant correlation between in vitro response and responses seenin patients. In addition, radiobiologic studies of the lung cancercell lines suggest both a greaterfractional cell kill of SCLClines,and significant heterogeneity of cell kill by radiation between the lines. Takentogether theseresultsprovide a beginning data basefor usingthis lungcancer cell linepanelto test newtherapies and combined modality therapy in vitro and suggestthatindividual patient's tumors varyin theirchemoand radiosensitivity. Molecularand cell biologic studiesof these same cell tines have also suggesteda working model for the pathogenesis of lung cancer. This model includes: 1.) carcinogen exposure throughcigarette smoking, and possibly through industrial and radon gas exposure which may be enhanced by inheritance of a certain debrisoquine metabolic phenotype (datafrom pharmacogenetic studies of others) (inherited predisposition); 2.) production of gastrin releasing peptide (GRP) and other regulatory peptides (including insulin and transferrin-like growth factors) by neuroendocrine cellsof the lungpotentially increased by exposure to factors in cigarette smoke; 3.) autocrine growth stimulation of these neuroendocrine cells and probable paracrine stimulation of other bronchial epithelialcells withresulting proliferation and increased growth factorproduction resulting in a largepopulation of cells ableto replicate in the bronchial epithelium ("Tumor Promotion") as well as highlevel expression of the transcription factor/proto-oncogene c-jun-A (AP-l family) found

112

Radiation Oncology, Biology, Physics

Volume 15, Supplement I, 1988

bothin nonnallung andmany lungcancerlineswhich mediates theeffectof tumorpromoters on increasing transcription of genes; 4.) accumulation of genetic lesions in the replicating bronchial epithelial cells from carcinogen exposure including deletions and translocations involving chromosome regions 3p(14.23), and the rb gene locus ("Initiation"); 5.) constitutive activation of myc family proto-oncogenes including e-myc, N-myc, and L.myc by several mechanisms including transcriptional activation, and loss of attenuation; 6.) addition of other genetic changes such as ras family gene mutations (which couldpotentiate and/orbypass autocrlne growth factoreffects), further activation of myc family members by generearrangements and amplification to giveprogressive andmetastatic lesions. This model focuses attention on: 1.) the correct sequential timingof somatically inherited geneticalterations in cellular protooncogenes and genes involved in autocrine growth regulation; 2.) the molecular genetics of autocrine growth factorproduction and/or activation of their signal transduction pathways; 3.) the identification of gene(s) involved in chromosome deletion syndromes suchas that at 3p(l4-23) and studyof theirmechanism of actionand possible inheritance in patientpedigrees; 4.) the structure, function, andmechanism of activated expression of the myc family of oncogenes in lungcancer; and5.) inherited genetic changes in patientpedigrees for carcinogen metabolism and thepotential for inherited recessive genesidentified bychromosomal deletions.

FRIDAY, OCTOBER 14, 1988 TOLERANCE OF THE CENTRAL NERVOUS SYSTEM (CNS) BIOLOGICAL BASIS, CLINICAL APPLICATION AND LEGAL IMPLICATIONS. E. van der Schueren*, W. Van den Bogaert*, A.J. van der Koge1** Department of Radiotherapy, University Hospital, Leuven (Belgium). Department of Radiotherapy, St. Radboud Hospital, Nijmegen (The Netherlands). While a large amount of data has accumulated on CNS tolerance to irradiation, it is striking to see that diverging and even conflicting guidelines are still offered for clinical practice. Several reasons for this can be distinguished. Firstly, the available information consists of a mixture of experimental and clinical data, the interpolation of which is not always easy. While it is usually difficult to convince clinicians of the usefulness of experimental results, too often such attitude is followed by misinterpretat-ion of data. A critical analysis of the relevance and the limitations has to be made. The most solid information obtained until now is the relation between fraction size and tolerance. This has been clearly demonstrated in the rat spinal cord model and indeed been confirmed by clinical experience. Even for this well documented fact, several aspects have to be clearly realised. First of all hard data for fraction sizes below 2Gy are still scarce and secondly biological endpoints (ED50, paralysis in 50% of the animals) are different from clinical tolerance (EDO-ED5). With changing slopes of dose-effect curves with increasing fractionation predictions based on 50% incidence levels might be diverging for a 5% or lower level. While experimental data seem to have a good predictive value for relative changes in tolerance as a function of fraction sizes, they provide no information on absolute tolerance in vi el'l of the differences in sensitivity between rodents and men. Also for kinetics of repair in MFD treatments, exstensive data is available on rates and half-times of repair for fraction sizes down to 4Gy. Still no data is available for 2Gy fractions as usually employed in clinical practice for which estimation have to be used. Finally for a number of factors influencing tolerance in clinical practice, only limited biological data are available. The most important. of these are age, site, volume and residual damage. The gradual development of more detailed insights on tolerance has bee n stimulated by the interaction between biologists and clinicians. Further progress requires continued exchange of information, awareness of limitations and a clear assessment of the remaining questions.