Human cancer cell lines (Hs578T, MDA-MB-468, HCT116, SW620 or HT29) were fluorescently labeled with DiI (red) and injected into the perivitelline space (PVS) of 2 days post fertilization (dpf) Tg(fli1:eGFP) zebrafish larvae (a–e). At 4 days post injection (dpi), zebrafish xenografts were evaluated regarding: apoptotic index—% of activated Caspase3 (f–j, u), mitotic index—% mitotic figures (f’–j’, v), tumor size (f–j, w), angiogenic capacity (k–o, x) and metastatic potential (p–t, y). White arrowheads indicate mitotic figures. Apoptotic index (u, ****P < 0.0001), mitotic figures (v, Hs578T versus MDA-MB-468 ****P < 0.0001, ***P = 0.0002, **P = 0.0049), tumor size (w, ****P < 0.0001, **P = 0.0065, *P = 0.0129), total vessel density (x, ****P < 0.0001, **P = 0. 0045) and metastatic potential (y, ****P < 0.0001, HCT116 versus SW620 **P = 0.0036, SW620 versus HT29 **P = 0.0048, Fisher’s exact test) are expressed as AVG ± SEM. The number of xenografts analyzed are indicated in the representative images and each dot represents one zebrafish xenograft. Results are from 3 (Hs578T and MDA-MB-468) and 2 (HCT116, SW620 and HT29) independent experiments, which are highlighted in different colors corresponding to each individual experiment. Statistical analysis was performed using an unpaired t-test or Fisher’s exact test. Statistical results: (ns) > 0.05, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Scale bars represent 50 μm. All images are anterior to the left, posterior to right, dorsal up and ventral down.

Human cancer cell lines (Hs578T, MDA-MB-468, HCT116, SW620 or HT29) were injected into the PVS of 2 dpf Tg(fli1:eGFP) zebrafish larvae. Zebrafish xenografts were treated in vivo with bevacizumab and compared with untreated controls. At 4 dpi, zebrafish xenografts were imaged by confocal microscopy (a–e’). The percentage of mitotic figures (f), apoptosis (g, ****P < 0.0001) and tumor size (h, ****P < 0.0001) were quantified. The outcomes are expressed as AVG ± SEM. The number of xenografts analyzed are indicated in the representative images and each dot represents one zebrafish xenograft. Results are from 3 (Hs578T and MDA-MB-468) and 2 (HCT116, SW620 and HT29) independent experiments, which are highlighted in different colors corresponding to each individual experiment. Statistical analysis was performed using an unpaired t-test. Statistical results: (ns) > 0.05, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Scale bars represent 50 μm. All images are anterior to the left, posterior to right, dorsal up and ventral down.

Human cancer cell lines (Hs578T, MDA-MB-468, HCT116, SW620 or HT29) were fluorescently labeled with DiI (in red) and injected into the PVS of 2 dpf Tg(fli1:eGFP) zebrafish larvae. Zebrafish xenografts were treated in vivo with bevacizumab and compared with untreated controls. At 4 dpi, zebrafish xenografts vasculature was imaged by confocal microscopy (max Z-projections) (a–e’). Total vessel density (f, **P = 0.0023) and vessel infiltration (g, **P = 0.0027) were quantified. Filament analysis was performed in Hs578T tumor-related vasculature. Hs578T xenografts confocal images of untreated and Bevacizumab-treated (h, h’) were used to perform 3D projections using Imaris (i–i’) and skeletonized images on ImageJ (j–j’). Number of branching points (k, **P = 0.007) and average vessel length (l, P = 0.61) were quantified by filament analysis. To analyze the functionality of Hs578T tumor-related vessels, tumor cells were labeled with DeepRed (in gray) to generate xenografts in Tg(fli1:eGFP; gata1:DsRed) zebrafish larvae. Xenografts were treated in vivo with bevacizumab and compared with untreated controls. At 4 dpi, zebrafish xenografts were mounted in low melting agarose to be visualized by live-imaging confocal microscopy (m–n’). The percentage of xenografts with erythrocytes inside blood vessels was quantified (o). All outcomes are expressed as AVG ± SEM. The number of xenografts analyzed are indicated in the representative images and each dot represents one zebrafish xenograft. Results are from 2 independent experiments, which are highlighted in different colors corresponding to each individual experiment. Statistical analysis was performed using an unpaired t-test or Fisher’s exact test. Statistical results: (ns) > 0.05, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Scale bars represent 50 μm. All images are anterior to the left, posterior to right, dorsal up and ventral down.

Representative image of an MDA-MB-468 zebrafish xenograft with a tumor in the PVS and several micrometastasis spread throughout the zebrafish larvae body, namely in brain, eye, gills and CHT (a). Human cancer cell lines (Hs578T, MDA-MB-468, HCT116, SW620 or HT29) were injected into the PVS of 2 dpf Tg(fli1:eGFP) zebrafish larvae. Zebrafish xenografts were treated in vivo with bevacizumab and compared with untreated controls. At 4 dpi, zebrafish xenografts were imaged by a fluorescent stereoscope to detect tumor cells throughout the zebrafish body (b). The percentage of xenografts that display micrometastasis in the CHT region was quantified (b, Hs578T **P = 0.0042, MDA-MB-468 **P = 0.0017, SW620 *P = 0.0112) and the outcomes are expressed as AVG ± SEM. Results are from 5 (Hs578T and SW620), 3 (MDA-MB-468) and 2 (HCT116 and HT29) independent experiments. The presence of micrometastasis in other organs besides the CHT was also quantified, namely in the brain, eye and gills, in untreated and bevacizumab-treated xenografts and the incidence of micrometastasis in two or more metastatic sites was determined (c, MDA-MB-468 from ~43.7 to ~23.4%, **P = 0.0081; HCT116 from ~19 to ~2.6%, *P = 0.0308; and HT29 from ~44 to ~20%, **P = 0.0061). Statistical analysis was performed using a Fisher’s exact test. Statistical results: (ns) > 0.05, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.

Schematic representation to distinguish between early and late metastatic steps (a). Injected xenografts were sorted into two groups—xenografts with only cells in the PVS and xenografts with cells in circulation. MDA-MB-468 and SW620 micrometastasis in the CHT region were quantified from both groups (b, MDA-MB-468 *P = 0.0246 and SW620 P = 0.06). Results are from 2 independent experiments and outcomes are expressed as AVG ± SEM. The number of xenografts analyzed are indicated  below the graphs. Statistical analysis was performed using a Fisher’s exact test. Statistical results: (ns) > 0.05, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.

Human breast cancer or CRC surgical resected samples were injected into the PVS of 2 dpf Tg(fli1:eGFP) zebrafish larvae. zPDXs were treated in vivo with bevacizumab and compared with untreated controls. At 4 dpi, zebrafish xenografts were imaged by confocal microscopy. The percentage of apoptosis (a, zPDX#3 *P = 0.0159, g = 1.75 and zPDX#5 *P = 0.0101, g = 1.51) and tumor size (b, *P = 0.0381, g = 1.08) were quantified. In parallel, zebrafish xenografts were analyzed in a fluorescent stereoscope to detect micrometastasis in the CHT region, followed by confocal confirmation (c). The outcomes are expressed as AVG ± SEM. The number of xenografts analyzed are indicated below the graphs. Results are from 1 independent experiment. Statistical analysis was performed using an unpaired t-test for apoptosis and tumor size and a Fisher’s exact test for micrometastasis. Statistical results: (ns) > 0.05, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Cohen’s D 1988 scale of effect size with Hedges’ g correction (g): g = 0,2 low; g = 0.5 moderate; g = 0.8 high.

Computerized tomography scans of Patient#7 (a–c’) and Patient#8 (h–j’) of different regions, including liver (a, a’, i, i’), lung (h, h’), mesenteric region (b, b’) and abdominal wall (c, c’, j, j’) in both pre- and post-treatment settings. Red arrows highlight metastasis. Human patient surgical/biopsy samples were injected into the PVS of 2 dpf Tg(fli1:eGFP) zebrafish larvae. zPDXs were treated in vivo with bevacizumab and compared with untreated controls. At 4 dpi, zebrafish xenografts were imaged by confocal microscopy (d–e’, k–l’). Activated caspase3 was quantified in both groups (f, zPDX#7 P = 0.4682 and g = 0.29; m, zPDX#8 P = 0.1817 and g = 0.82). In parallel, the percentage of xenografts that display micrometastasis in the CHT region was quantified (g, zPDX#7 P = 0.73 and g = 0.11; n, zPDX#8 P = 0.11 and g = 0.42). The outcomes are expressed as AVG ± SEM. Number of zPDX analyzed for each condition is indicated in the figure. Results are from 1 independent experiment. Statistical analysis was performed using an unpaired t-test for apoptosis and a Fisher’s exact test for micrometastasis. Statistical results: (ns) > 0.05, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Cohen’s D 1988 scale of effect size with Hedges’ g correction (g): g = 0,2 low; g = 0.5 moderate; g = 0.8 high. Scale bars represent 50 μm. All images are anterior to the left, posterior to right, dorsal up, and ventral down.