A phase I trial of isolated hepatic perfusion (IHP) using 5-FU and oxaliplatin in patients with unresectable isolated liver metastases from colorectal cancer.

BACKGROUND: Isolated hepatic perfusion (IHP) with melphalan is an established approach for patients with unresectable metastatic liver lesions. This study determined the safety and maximum tolerated dose (MTD) of 5-FU with oxaliplatin via IHP. METHODS: Standard 3 × 3 Phase I design. Subjects with unresectable isolated CRC liver metastases scheduled for HAI pump were eligible. IHP used fixed-dose oxaliplatin with escalating 5-FU doses. Toxicity (CTCAE v 4.0) and response (RECIST), progression-free survival, and overall survival (OS) were assessed. Systemic and IHP plasma PK of 5-FU, anabolites, and platinum were determined. RESULTS: All 12 patients had received ≥ 1 line of pre-IHP chemotherapy. There were 4 grade 3 serious adverse events (33.3 %) and 1 grade 4 event (8.3 %). Also, 2 dose-limiting toxicities occurred at DL2 at 300 mg/m(2), resulting in expansion of DL1 at 200 mg/m(2) 5-FU, the eventual MTD. At 6-month follow-up, 9 patients (82 %) demonstrated partial response, while 2 (18 %) exhibited stable disease. Also, 64 % of patients demonstrated a >50 % decrease in CEA. The 1- and 2-year OS probabilities were 90.9 and 71.6 %, respectively, with median follow-up of 24 months. IHP exposures (AUC0-60 min) were 10.9 ± 4.5 µgPt h/mL, 49.3 ± 30.7 µg h/mL 5-FU (DL1), and 70.5 ± 35.5 µg h/mL 5-FU (DL2). Systemic exposure (AUC0-inf) relative to IHP exposure was negligible for both platinum (1.1 ± 1.5 %) and 5-FU (0.09 ± 0.10 %). CONCLUSIONS: The MTD for IHP was 200 mg/m(2) 5-FU with 40 mg/m(2) oxaliplatin. Systemic exposure to the agents was minimal during IHP. The response and survival observed warrants assessment in a larger phase II trial.

Tumor progression, micrometastasis, and genetic instability tracked with histochemical marker genes.

Mouse fibrosarcoma (3T3 cells transfected with different oncogenes), human neuroblastoma, or human prostate carcinoma cells have been genetically-tagged with different histochemical marker genes (E. coli lacZ, placental alkaline phosphatase, or Drosophila alcohol dehydrogenase). Injection into athymic nude mice permits their tracking at all stages of primary tumor formation and micrometastasis to various organs at the single-cell level. Two different tumor classes, tagged with different marker genes, can be tracked together. Primary tumors display regional dominance of one tumor class with exclusion of other classes. During micrometastasis, tumor cells are detected binding to the endothelium of lung blood vessels, followed by establishment of multiple-cell micrometastases. Micrometastases in some organs are transient while in other organs there is differential expansion into overt metastases. Tagged tumors also reveal the timing of angiogenesis of developing primary tumors and overt metastases. In all three tumor systems, there are three classes of genetic stability of marker gene expression in clonal populations-high stability, intermediate stability, and high instability. Instability in marker gene expression in one tagged prostate carcinoma system does not depend on a hypermethylation mechanism, suggesting a genetic basis for loss of activity. Use of histochemical marker genes, combined with laser-capture microdissection and various PCR methods, can now be used to evaluate gene activities in single or multiple tumor cells in virtually any organ and primary tumor of the animal model system.

Tracking micrometastasis to multiple organs with lacZ-tagged CWR22R prostate carcinoma cells.

Metastasis to organs other than lung is rarely observed in animal model systems of human prostate carcinoma (PCA), with the exception of already metastatic isolates of human PCA cultured for long periods of time. To analyze more directly the evolution of metastatic variants from primary PCA tumor isolates, the lacZ histochemical marker gene was transfected into the CWR22Rv1 cell line isolated from the CWR22R xenograft (primary tumor). Three clones of varying lacZ-expression stability were analyzed for tumorigenicity and progression in athymic nude mice. Clones B and D were highly tumorigenic in the subcutis; however, lacZ expression was highly unstable. In contrast, clone H demonstrated highly stable lacZ expression for >25 passages in culture or in animals. Clone H, injected sc in a PBS vehicle, gave a 15-40% tumorigenic take. All primary tumor-bearing animals exhibited micrometastases in lung and other organs. Clone H injected in a Matrigel vehicle gave 100% tumorigenicity, with all animals displaying micrometastases in lung, liver, and/or bone (lower frequency in brain and kidney). Overall, the relative frequency of micrometastasis to multiple organs was lung>liver=bone>>brain>kidney. Overt metastases were never observed in the lung or bone but were occasionally found in liver. lacZ-transfected clone H CWR22Rv1 cells represent a much more accurate model of metastasis of PCA to the organs normally involved in progression of the human disease. Use of marker gene-tagged cells and other high-resolution molecular techniques will now permit analyses of the earliest events in PCA progression and micrometastasis.

Tracking prostate carcinoma micrometastasis to multiple organs using histochemical marker genes and novel cell systems.

Studies of human prostate carcinoma (PCA) have been hampered by only a few cell systems from already-metastatic human disease. We have developed a novel cell system by using tissue cultured CWR22R cells from a xenograft of a primary tumor from a human patient. These cells were transfected with the bacterial lacZ gene to maximize their detection during progression and metastasis in nude mice. LZ-CWR22R cells are extremely stable for lacZ expression over 25 passages and metastasize to lung, liver, and bone from the subcutis - major sites of metastasis of the human disease. A matrigel vehicle facilitated development of primary tumors and micrometastases in all organs. While some micrometastases developed into overt metastases, others remained as micrometastases for long periods of time, possibly providing a model of latency of metastatic disease. An experimental metastasis model (tail vein injection) also generated micrometastases in lung, liver, and bone with differing kinetics of formation and stability. Serial sections of many individual lung micrometastases within one hour of injection indicated considerable heterogeneity in cellular composition (from 1 to 19 cells/site) while liver sites at later times were comprised of only 1 or 2 cells (the size of bone sites were comparable to those of liver). By combining use of these histochemically-tagged PCA cell systems with high resolution molecular analyses (laser-capture microdissection), it will now be possible to analyze gene expression patterns characteristic of micrometastases developing in several different organs.