Allan J. Franko

Binding of 3H-misonidazole to solid human tumors as a measure of tumor hypoxia☆

Treatment-resistant, chronically hypoxic tumor cells have been assumed to exist in some solid human tumors, limiting their curability. To date, six patients with different types of tumors have been studied using radioactive labelled electron affinic compounds that bind to hypoxic cells. Although the gross clinical appearance of the tumors in all six patients was of a large and fixed mass which might on clinical grounds be expected to contain hypoxic cells, we have observed drug binding to hypoxic regions in only two, a rapidly growing small cell lung cancer (SCLC) and a malignant melanoma. The hypoxic fraction of the malignant melanoma was found to be 6% and the SCLC tumor approximately 10%. We have observed that areas of maximum adduct formation can be found in tumor cells immediately adjacent to blood vessels, suggesting that blood flow over the labelling interval was restricted. These preliminary studies suggest that sensitizer adduct formation in human tumor tissue may be a useful measure of tissue pO2 at the cellular level and that tumor hypoxia might be more related to the rate of tumor growth and histological grading than to tumor size.

A quantitative histological study of strain-dependent differences in the effects of irradiation on mouse lung during the intermediate and late phases.

Strain differences in the intermediate and late phases of the radiation response of mouse lung were investigated histologically. The proportion of lung impairment in mice at 28 and 52 weeks postirradiation and in mice dying of respiratory insufficiency was assessed by scoring lung acini as nonfunctional due to lesions which obstructed airflow, or open and presumably functional. The nine strains tested were divided into three groups on the basis of the late fibrotic response. Group 1 mice, three C57 strains, developed extensive contracted fibrosis and usually showed enough damage to explain late deaths. Group 2, SWR, A, and BALB/c strains, developed foci of contracted fibrosis. Group 3, CBA and two C3H strains, did not form fibrotic scars. Mice in Groups 2 and 3 that died with no pleural effusions appeared to have insufficient late lung damage to account for respiratory distress. Problems with pulmonary blood flow were indicated by evidence of loss of fine vasculature and right ventricular hypertrophy. In nondistressed, late-stage mice in Groups 2 and 3, loss of capillary perfusion in lung parenchyma free of obvious lesions was demonstrated by infusion of colloidal carbon. In one strain, A, an estimate of the proportion of nonperfused lung was made on distressed late-stage mice. Almost 50% of lung acini were nonfunctional as a result of nonperfusion, and an additional 9% of acini were nonfunctional due to lesions obstructing ventilation. It is suggested that nonperfusion of apparently normal lung acini is a major factor in late-phase deaths in those mouse strains which show little or no fibrosis.

Immunohistochemical Localization of Transforming Growth Factor β and Tumor Necrosis Factor α in the Lungs of Fibrosis-Prone and "Non-Fibrosing" Mice during the Latent Period and Early Phase after Irradiation

To evaluate the possibility that TGF-beta and TNF-alpha are involved in fibrosis induced in mouse lung by irradiation, the proportion of cells immunoreactive for each was compared in two strains of mice. C3HeB/FeJ mice develop only classical pneumonitis during the early phase, whereas C57L/J mice develop small, tightly packed areas of inflammation which undergo fibrosis during the latent period, and exhibit progressive fibrosis of large regions of intense inflammation during the early phase. Very few cells were immunoreactive for an antibody to the latency-associated peptide (LAP) of TGF-beta during the latent period in C3HeB/FeJ mice, and no cells were positive during the early phase. In contrast, between 0.7 and 10% of cells were positive in C57L/J mice in lesions without fibrosis and in lesions in the early stages of fibrosis. Fibroblasts positive for LAP were seen only in lesions containing fibrosis. A similar pattern of immunoreactivity was seen in C57L/J mice using an antibody which recognizes active TGF-beta, with the exception that positive fibroblasts were observed within areas of inflammation without fibrosis. Thus the association of active TGF-beta with fibroblasts might be a characteristic of the initiation of fibrosis in this model. TNF-alpha was detected in macrophages in all classes of lesions, and minor differences between the strains did not appear to be biologically meaningful.

A quantitative histological study of strain-dependent differences in the effects of irradiation on mouse lung during the early phase

Strain differences in the radiation response of mouse lung during the early phase (before 28 weeks postirradiation) were investigated histologically. The nine strains tested were divided into three groups on the basis of the nature of the edema present, the occurrence of hyaline membranes, and the presence of fibrosis. Group 1 mice, three C57 strains, developed hyaline membranes, focal fibrosis, and a protein-rich edema containing fibrin. Group 3, CBA and two C3H strains, had only a protein-poor edema with little fibrin and developed no visible fibrosis. Group 2 mice had both types of edema and small quantities of focal fibrosis. The degree of lung impairment in mice dying of respiratory insufficiency was assessed by scoring lung acini as nonfunctional or open and presumably functional. Over 70% of acini were nonfunctional as a result of airflow obstruction. This was considered sufficient to account for death. Carbon perfusion immediately before sacrifice indicated that all types of lesions were at least partially perfused with blood. Pleural effusions were found in some individuals of two strains. The proportion of nonfunctional acini was similar in mice of the same strain with and without effusions, which would not be expected if the effusions contributed appreciably to respiratory distress in the early phase.

Misonidazole and other hypoxia markers: Metabolism and applications

A substantial effort is being devoted to developing markers for hypoxia in tumors. Most of the work to date has been performed on misonidazole (MISO), which is selectively metabolized by hypoxic cells to reactive products that bind covalently to cellular constituents. This paper attempts to review the metabolism of MISO as it relates to binding, to summarize several of the properties of the binding of MISO to cells and tissues which appear to be directly relevant to the characteristics of the reactive species involved, and to evaluate the potential of MISO and other nitroheterocycles as markers for hypoxia. Four roles for a hypoxic marker are considered. MISO labeled with 3H or 14C is a good marker for local radiobiological hypoxia in autoradiograms of tumor sections, but more work is required to investigate factors other than oxygen concentration that conceivably might affect the binding process. In quantitating hypoxic fraction in tumors using non-destructive techniques, which has been modelled by correlating surviving fraction with 14C-misonidazole uptake, non-specific binding to aerobic and necrotic tissue limits the accuracy of the estimate, but useful clinical applications can still be envisaged. For quantitation of a change in the hypoxic fraction of an individual tumor using serial assays, preliminary data suggest that MISO binding should be a sensitive assay. Fluorescent nitroheterocycles have a great deal of potential as markers to enable the sorting of tumor cell suspensions into portions derived from the hypoxic and aerobic regions, but better compounds are needed.

On the nature of the radiobiologically hypoxic fraction in tumors

Since the classical work of Thomlinson and a it is commonly assumed that chronically hypoxic cells have stopped cycling in the G,-phase of the cell cycle, a phase which has often (though not always) been shown to be inherently more radiosensitive under aerobic conditions than the cells in S-phase. Thus. when hypoxia is imposed upon the cycling cells by nitrogen it would be expected that in many cases the survival curve slope would reflect greater resistance than G,-phase chronically hypoxic cells. This is not generally observed and can be explained several ways: I) chronically hypoxic cells are not all in the G,-phase, 2) chronically hypoxic cells in the G,-phase are different in terms of radio-sensitivity (i.e.. more radioresistant than G,-phase cells in culture which have been made hypoxic acutely), 3) the hypoxic fraction in tumors in air-breathing animals contains not only chronically hypoxic cells, but also acutely hypoxic cells in all parts of the cell cycle (including S-phase cells) as a result of abrupt changes in vascular perfusion. In this case. a large proportion of the cells which are hypoxic during normal air-breathing would be similar in cell cycle prop-

Development of Fibrosis after Lung Irradiation in Relation to Inflammation and Lung Function in a Mouse Strain Prone to Fibrosis

The development of lung fibrosis after single-dose thoracic irradiation was studied histologically in C57L/J male mice. Lung function was monitored using uptake of carbon monoxide. During the latent period (prior to 15 weeks postirradiation) mice were chosen at random, while during the early phase (15-22 weeks) mice were sacrificed when they developed a functional deficit of at least 50%. Excess mice with a deficit of 50% in the early phase were followed into the late phase and sacrificed at 31 or 40 weeks. Two scoring methods were used to quantify lung damage. Fibrotic lesions and foci of inflammation were counted for the latent period, and the proportion of nonfunctional acini was determined for the early and late phases. After a dose 1 Gy less than the LD50/180, small regions of mild inflammatory infiltration appeared at 6 weeks, and small, focal fibrotic lesions containing numerous macrophages appeared at 8 weeks postirradiation. The number of fibrotic lesions increased steadily during the latent period in a manner that is consistent with conversion of inflammatory lesions to foci of fibrosis. Mice sacrificed upon developing a 50% functional deficit during the early phase had approximately equal proportions of lung affected by fibrosis and inflammation. Those mice which developed a similar respiratory deficit in the early phase and were followed into the late phase usually showed little change in lung function. However, when sacrificed at 31 weeks they had twice as much fibrosis and very little inflammation, suggesting that the inflammatory lesions had become fibrosed. The average number of macrophages per unit area of fibrosis declined during the latent period and changed little during the early and late phases. Lymphocytes and mast cells were also quantified in fibrosed regions.

Covalent binding of a fluorinated 2-nitroimidazole to EMT-6 tumors in Balb/C mice: detection by F-19 nuclear magnetic resonance at 2.35 T.

A hexafluorinated 2-nitroimidazole (CCI-103F) has been synthesized and its properties as a hypoxic cell binder investigated. The drug has a plasma half-life of 90 min in Balb/C mice and a tumor-to-plasma ratio of 0.8. Following sustained exposure to the drug and a washout period for the unbound drug, liver and tumor samples were excised. NMR investigation of the excised tissue in a 2.35 T animal sized NMR facility revealed the presence of 19F bound to both tissues. Further improvement in sensitivity is required to make 19F NMR detection of binding in situ feasible.

The genetic basis of strain-dependent differences in the early phase of radiation injury in mouse lung.

Substantial differences between mouse strains have been reported in the lesions present in the lung during the early phase of radiation injury. Some strains show only classical pneumonitis, while other strains develop substantial fibrosis and hyaline membranes which contribute appreciably to respiratory insufficiency, in addition to pneumonitis. Other strains are intermediate between these extremes. These differences correlate with intrinsic differences in activities of lung plasminogen activator and angiotensin converting enzyme. The genetic basis of these differences was assessed by examining histologically the early reaction in lungs of seven murine hybrids available commercially after whole-thorax irradiation. Crosses between fibrosing and nonfibrosing parents were uniformly nonfibrosing, and crosses between fibrosing and intermediate parents were uniformly intermediate. No evidence of sex linkage was seen. Thus the phenotype in which fibrosis is found is controlled by autosomal recessive determinants. Strains prone to radiation-induced pulmonary fibrosis and hyaline membranes exhibited intrinsically lower activities of lung plasminogen activator and angiotensin converting enzyme than either the nonfibrosing strains or the nonfibrosing hybrid crosses. The median time of death of the hybrids was genetically determined primarily by the longest-lived parent regardless of the types of lesions expressed.

Rate of Death of Hypoxic Cells in Multicell Spheroids

The rate of death of hypoxic cells was measured in multicell spheroids, which are considered to model in vitro the microenvironments surrounding such cells within solid tumors. Two types of experim...