William G. Hammond

Successful xenotransplantation of human lung cancer correlates with the metastatic phenotype

BACKGROUND Occult micrometastases could explain deaths from stage I non-small cell lung cancer (NSCLC) after complete resections. If patients who have occult metastases could be identified, systemic therapy might be beneficial. METHODS Non-small cell lung cancers from 81 patients in stages I, II, and III were transplanted to nude beige mice. Mean follow-up was 22.5 months (2 to 61 months). RESULTS Twenty-one xenotransplants successfully took, and seven metastasized in the nude mice. Neither the predominant cell type nor the incidence of lymph node metastases correlated with the results of xenotransplantation. Of the 21 patients whose NSCLCs took in xenotransplantation, 13 (61.9%) have had development of metastases, and 9 (42.9%) have died of the cancer. Among the 57 patients whose NSCLCs did not take, 14 (24.6%) have had development of metastases, and 9 (15.8%) have died of their cancer. The higher incidence of metastases in association with xenotransplant take is significant (p = 0.0032). CONCLUSIONS Patients whose NSCLCs take in xenotransplantation are at high risk for metastases. The xenotransplantation model is a step toward facilitating precise cellular biologic definition of the metastatic propensity of human NSCLC:

A Clinically Relevant Canine Lung Cancer Model

Research on early human lung cancer is difficult; we have sought a canine correlate. Regimens included endobronchial submucosal injections and topical focal applications of benzo[a]pyrene, nitrosomethylurea, dimethylbenzanthracene, and methylcholanthrene, singly or in combinations. Sustained-release discs were placed into lung parenchyma or sutured into major bronchi. Tracheal segments were isolated as cervical pedicle grafts. Gross and histological evolution was reproducible. Columnar and basal hyperplasia and squamous metaplasia were early changes. Atypia occurred within 6 weeks and was found in all dogs within 16 to 18 weeks. Invasive cancers occurred within 8 to 65 months. No tracheal graft developed cancer. Of 15 dogs with parenchymal sustained-release implants, 1 to date has developed cancer in 8 months. Four endobronchial regimens have produced 16 cancers in 56 lungs at risk for 18 to 65 months. No cancers developed in 23 lungs at risk from eight other regimens. Of 10 dogs at risk for unilateral endobronchial cancer, 5 have had cancer. Of 23 dogs with both lungs at risk, 9 developed cancer. We have shown focal carcinogenesis with well-defined pathogenesis and an extended preneoplastic period at predictable sites in a lung cancer model.

Bronchiolo-alveolar regions in adenocarcinoma arising from canine segmental bronchus

Adenocarcinomas induced in canine bronchial segments placed subcutaneously have bronchiolo-alveolar regions. Immunocytochemistry and routine staining of adjacent sections strongly suggests that the lining of these regions consists of type II cells. These regions may thus represent true prospective alveolar regions, as also seen in embryonic lungs. This first observation of bronchoalveolar cancer arising from a major bronchus indicates that the carcinogen-induced neoplastic progression in bronchial epithelium may lead to type II cell differentiation and type II cell tumor development. The preservation of cell properties in serial nude mouse transplants suggests that it is a stable type II cell population.

Alveolar stem cells in canine bronchial carcinogenesis

Alveolar type II cells are not present in normal epithelium of canine segmental bronchi but after carcinogen exposure they do occur in intra-epithelial lesions with all degrees of atypia and in invasive lesions with different glandular growth patterns. Immunohistochemistry for proliferation markers (PCNA; Ki-67) strongly suggest that such novel type II cells are pluripotential stem cells in canine bronchial carcinogenesis. Very likely, bronchial carcinogenesis is subject to an oncofetal mechanism of differentiation: bronchial epithelial retrodifferentiation followed by novel differentiation of alveolar tumor stem cells.

Lung cancer model for study of the metastatic process

In our hamster model of focal, chemically induced nonsmall cell lung cancer (NSCLC), we studied metastases in autochthonous hamster hosts (n = 300) and in syngeneic hamster and nude mice recipients (n = 230) of serial tumor transplants. Metastases in autochthonous hosts and transplant recipients occurred in regional lymph nodes, liver, and adrenals. In autochthonous host hamsters no metastases were noted from microinvasive (n = 112) or visible cancer less than 3.0 mm in diameter (n = 66); the incidence of metastasis was 8.2% (4/49) from 3- to 10-mm cancers and 22% (16/73) from cancers 10 mm in diameter or larger (p less than 0.05). Serial transplants were used to evaluate the metastatic propensity of 20 primary and six metastatic NSCLCs. Six primary NSCLCs that metastasized in the autochthonous host and six metastatic NSCLCs all metastasized promptly in recipients. This expression of metastatic potential was significantly different (p less than 0.05) from 14 primary cancers without autochthonous host metastases. Eight of the 14 caused no metastases in recipients, even after 5 to 11 tumor growth cycles; metastases occurred from the other six primary NSCLC after 3 to 12 tumor growth cycles in transplant recipients. Primary hamster NSCLCs metastasize in the autochthonous host with a frequency and a distribution pattern similar to human NSCLCs. A new model to study serially the cellular changes that govern the process of metastasis in NSCLC has been developed.

Logistic Regression Analysis in Experimental Bronchial Carcinogenesis

Abstract The technique of logistic regression is shown to be a usable method for testing the significance of associations in the two-sample problem of comparing rates of progression through an ordered series of states as a function of another ordered variable (such as time). The technique is illustrated using an example comparing the development of preneoplastic and neoplastic changes in two groups of mice exposed to a carcinogen and subsequently receiving or not receiving a potential modifier of carcinogenesis.

Metastases from fresh human non-small cell lung cancers propagated in nude mice

Specimens from 69 freshly resected human non-small cell lung cancers (NSCLC) were transplanted into nude mice. Twelve mice died before the transplants were evaluable. There were 4 takes of 12 evaluable transplants into untreated athymic nude mice and 24 takes of 45 evaluable transplants into nude mice with decreased natural killer (NK) cell activity. Fourteen tumor lines were propagated into 2 or more successive transplant generations. Distant metastases occurred from the mid-dorsal transplant site in 8 of 81 (9.88%) recipients of 4 of those tumor lines, after 3-9 consecutive tumor growth cycles. Xenotransplantation of freshly resected human NSCLC provides a model with potential for serial assessment of cellular changes related to metastatic capability.

Metastatic behaviour of canine lung carcinoma in autochthonous and xenotransplant hosts

Specimens from 24 chemically induced canine non-small cell lung cancers (NSCLC) were xenotransplanted into nude mice. Twenty-one tumour lines were established in serial transplantation; four were from NSCLC that arose from orthotopically induced NSCLC in four dogs, and 17 were from NSCLC that arose heterotopically in 15 subcutaneous bronchial autografts (SBA) in seven dogs. Distant metastases developed in recipients of two orthotopic NSCLC after three and eight consecutive tumour growth cycles; no metastases have occurred after three and six growth cycles of two other orthotopic tumour lines. Recipients of eight heterotopic tumour lines developed metastases after 3–9 consecutive tumour growth cycles, while no metastases have occurred after 4–11 growth cycles in recipients of nine other heterotopic tumour lines. In three instances, both metastasizing and non-metastasizing tumour lines resulted from NSCLC that arose in different SBAs in the same dog. These findings indicate that, in the canine SBA bronchogenic cancer model as expanded by tumour xenotransplantation, those molecular events involved in both the initiation and the full progression of a single cancer may be investigated serially and concomitantly.

Histogenesis of adenosquamous bronchogenic carcinoma

In our hamster lung cancer model studies, among 463 non-small-cell lung cancers (NSCLC), there were 47 adenosquamous neoplasms. In 24 of 27 lesions with diameters of less than 3.0 mm, the adenocarcinoma and the squamous cell carcinoma components arose as separate, spatially discrete lesions, but these were separate in only 7 of 20 lesions with diameters of 30 mm or greater. Co-infiltration of the components became more frequent as tumor size increased. The usual adenosquamous variety of NSCLC is likely a collision tumor, with each component possessing separate biological characteristics. Thus, future prognostically directed studies of this variety of NSCLC must recognize that these neoplasms have two components, each of which needs to assessed.

Loss of nm23 and Alu DNA in human lung cancer propagated in nude mice.

Freshly resected human non-small cell lung cancer (NSCLC) has been successfully transplanted and propagated subcutaneously in nude mice (Cancer Letters 61 (1991) 53-60). We used this model to study the changes of the human metastasis suppressor genes, nm23-H1 and nm23-H2, through the process of propagation and metastasis of human NSCLC. Using a non-radioisotopic Southern analysis, the nm23-H1 and nm23-H2 genes were detected without evidence of deletion in the early generations of the tumor grafts. These genes, however, were absent from the tumor grafts sampled past 4 generations of propagation and from all the propagated metastases originated from the subcutaneous grafts. Further restriction analysis revealed that only mouse DNA, but no human Alu DNA, was present in the tumor specimens which lacked the human nm23 genes. Thus, there is a loss of human DNA but a gain of mouse DNA in the propagated tumors originated from the transplanted human NSCLC. The mechanisms of loss of human DNA in these propagated tumors in nude mice have yet to be delineated.