Validation of an orthotopic model of human lung cancer with regional and systemic metastases.

BACKGROUND: We developed an orthotopic model of human lung cancer that exhibits highly predictable regional and systemic metastases. This study examines the response of the model when treated with conventional and experimental chemotherapy. METHODS: NCI-H460 tumor fragments were implanted into the right caudal lung lobe of a nude rat. Treatment commenced 2 weeks later. We assessed response by comparing primary tumor and mediastinal lymph node weights, total body weight, and length of survival with untreated, tumor-bearing control animals. We also calculated the incidence of metastasis to kidney, bone, brain, and contralateral lung in treated versus untreated animals. RESULTS: Mitomycin and cisplatin showed broad activity against primary and metastatic disease. The matrix metalloproteinase inhibitor batimastat, low-dose cisplatin, and mitomycin significantly prolonged survival. High-dose cisplatin caused renal toxicity that shortened survival. Brain metastases did not respond to mitomycin, consistent with its poor blood-brain barrier penetration. CONCLUSIONS: Responses were similar to NCI-H460 in vitro data and consistent with clinical experience for these drugs. Drug-related toxicities similar to those seen in clinical practice were detected.

Characterization of a highly metastatic, orthotopic lung cancer model in the nude rat.

The prevailing subcutaneous nude rodent tumor xenograft models used for biological and preclinical studies do not optimally reflect some important biological properties of cancer, especially invasion and metastasis. Orthotopic models have been developed to address this need. However, for lung cancer none of the available models are optimal, in that none originate from an orthotopic (bronchial) primary site and exhibit extensive extrathoracic metastasis. Our goal was to develop a consistent rodent model of non-small cell lung cancer with both of these properties. Groups of male Rowett nude rats were given 500 rads of gamma radiation and then endobronchially implanted in the right caudal lobe airway with 50 mg of small NCI-H460 tumor fragments taken from an orthotopic donor tumor. They were then sacrificed at selected post-implantation times and evaluated grossly and histologically for animal weight, primary tumor take and size, and metastatic tumor incidence at multiple sites. At a late time point (32-35 days), consistency of primary tumor size and metastasis was estimated by comparing results from four groups of rats implanted on different occasions. The results showed that the primary tumors grew steadily, reaching four grams by days 32-35. Rats gained weight until days 14 to 21, but then began to show cachexia. High metastatic rates (>60%) were seen for mediastinal lymph nodes (by 21 days), and kidney, bone and brain (by 28 days). Mean primary tumor size and the incidences of both regional and systemic metastasis were consistent at 32-35 days in four different groups of six animals. In conclusion, this orthotopic lung cancer model is highly metastatic and consistent in terms of both primary tumor growth and metastatic behavior. It is the only available rodent model of human lung cancer emanating from an endobronchial site and metastasizing to multiple extrapulmonary sites, and should be very useful for both biological and preclinical studies of lung cancer, particularly where studies of antimetastatic activity are of interest, and/or where survival studies are desired.

Time-related effects of enzymatic disaggregation on model human lung carcinomas.

Enzyme cocktails used to prepare tumor cell suspensions may influence yield, viability, and cytology, thus time-related cocktail effects on model human lung carcinomas were examined. A549, NCI-H125, and NCI-H460 carcinomas were completely disaggregated at 25 degrees C over 2 h with either (mg/ml) collagenase/DNAase (C/D, 1/0.1), collagenase/hyaluronidase/DNAse (C/H/D, 1/0, 1/0.1), or polymyxa protease/DNAse (PP/D, 3/0.1). Trypan blue viabilities, total yields, viable yields, and flow cytometric percent tumor cells (TC) were measured every 20-30 min (n = 4-7 per tumor type). The final percentages of TC, mononuclear cells (MN), polymorphonuclear cells (PMN), lymphocytes, and necrotic cells were determined by cytology (n = 4-5 per tumor type). The time-dependent measurements showed that 1) disaggregation was progressive and complete with all cocktails; 2) viability was stable or increasing with all cocktails; 3) percent TC was stable for all cocktails, but lower for PP/D than C/D in final suspensions; and 4) PP/D gave lower final total yields, higher final viabilities, but the same final viable yields as the C cocktails, suggesting selective elimination of dead cells by PP/D. Final cytology measurements showed that PP/D gave a lower percent MN and a higher percent PMN than C cocktails. Cocktail effects may importantly influence cell suspension properties.

Measuring gene mutation in vivo.

Both chromosomal aberrations and intragenic mutations are important in the activation of oncogenes and inactivation of antioncogenes. It is now possible to measure both types of mutational events in vivo in a single cell type. The results obtained in both Chinese hamsters and rats are surprising: mutations can be induced in the absence of detectable chromosomal aberrations by some agents and vice versa by others. Preliminary results from a validation study are presented.