Evidence for induction of a tumor metastasis-receptive microenvironment for ovarian cancer cells in bone marrow and other organs as an unwanted and underestimated side effect of chemotherapy/radiotherapy

Abstract

Background: One of side effects of chemotherapy and radiotherapy is the induction of several factors in various tissues and organs that create a pro-metastatic microenvironment for cancer cells that survive initial treatment.

Methods: In the present study, we employed human ovarian cancer cell line A2780 and immunodeficient mice xenograft model to test effect of both ibuprofen and dexamethasone to ameliorate the therapy-induced pro-metastatic microenvironment in bone marrow, liver, and lung.

Results: In our studies, we found that total body irradiation or administration of cisplatin increases the metastatic spread of human ovarian cancer cells transplanted into immunodeficient mice compared with animals unexposed to irradiation or cisplatin. Moreover, conditioned media harvested from irradiated murine bone marrow, lung, and liver chemoattracted human ovarian cancer cells, and this chemotactic activity was inactivated by heat, suggesting a major involvement of peptide or peptide-bound chemoattractants. We also observed that human ovarian cancer cells proliferate better if exposed to cell debris harvested from irradiated murine bone marrow. Finally, the pro-metastatic microenvironment in mice induced by radio- or chemotherapy was significantly ameliorated if animals were treated at the time of radiotherapy administration with non-steroid (ibuprofen) or steroid (prednisone) anti-inflammatory drugs.

Conclusions: In summary, we propose that a radiochemotherapy-induced, pro-metastatic microenvironment plays an important role in the metastasis of cancer cells that are resistant to treatment. Such cells have characteristics of cancer stem cells and are highly migratory, and simple, intensive, anti-inflammatory treatment by non-steroid agents to suppress induction of pro-metastatic factors after radiochemotherapy would be an interesting anti-metastatic treatment alternative.

Figure 1

Radiochemotherapy and anti-inflammatory drug injection strategy. SCID-Beige inbred mice received a chemodrug (cisplatin) for three consecutive days, −D1, D0, and D1, and A2780 cells were transplanted 24 hrs after the first cisplatin injection. Similarly, SCID-Beige inbred mice underwent irradiation at 1000 cGy on –D1, prior to A2780 cell injection. SCID-Beige inbred mice also received anti-inflamatory drugs (ibuprofen or dexamethasone) for five consecutive injections (−12 hrs, 0 hrs, +12 hrs, +24 hrs, +36 hrs) and underwent irradiation at 1000 cGy at 0 hrs, followed immediately by A2780 celltransplantation. All drugs were injected at doses explained in the figure. All mice were sacrificed 48 hours after the injection of A2780 cells, and bone marrows, spleens, livers, and lungs were collected. The presence of ovarian cells (i.e., murine–human chimeras) was evaluated by the level of human α-satellite DNA expression.

Figure 2

Chemotherapy and irradiation create a pro-metastatic microenvironment in various murine organs that chemoattract A2780 cells. qRT-PCR detection of A2780 human ovarian cancer cells in bone marrow (BM), liver, lung, and spleen after irradiation (1000 cGy) and chemotherapy (cisplatin). Five mice were employed per group. The results from two independent experiments are combined and shown as mean ± SD. *p < 0.05 or **p < 0.005 compared with control (organs neither irradiated nor received cisplatin).

Figure 3

Anti-inflammatory treatments inhibit the metastatic spread of A2780 cells. Ibuprofen (Panel a) and dexamethasone (Panel b) significantly decrease the metastatic spread of A2780 human cancer cells into bone marrow (BM), liver, lung, and spleen of SCID-Beige inbred mice. In these experiments, five mice were used per group, and the results from three independent experiments for each anti-inflammatory drug are shown as mean ± SD. *p < 0.05, **p < 0.005, or ***p < 0.0005 for the group receiving irradiation (1000 cGy) compared with the control group (no irradiation) and for the groups receiving irradiation and an anti-inflammatory drug compared with the group receiving irradiation alone.

Figure 4

Conditioned media (CM) derived from irradiated organs are potent chemoattractants for A2780 cells. a. Chemotaxis of A2780 cells across Transwell membranes in response to CM harvested from BM, liver, and lung of irradiated and ibuprofen-treated SCID-beige inbred mice. b. Chemotaxis of A2780 cells across Transwell membranes in response to irradiated BM CM samples subjected to heat inactivation at different temperatures (left panel) and derived from molecular filtration by using Centricon centrifugal filtration devices (right panel). The chemotaxis assay was performed at least three times in duplicate, with similar results. Results are presented as mean ± SD, with a statistical significance *p < 0.05 or **p < 0.005 relative to the control (cells stimulated with RPMI 0.5% BSA medium).

Figure 5

Damaged bone marrow cells and CM harvested from organs from irradiated mice support A2780 cell growth in vitro. a. Proliferation and expansion of A2780 cells co-cultured with irradiated BM cells from C57BL/6 GFP mice for 48 hours, showing that irradiation causes lethally damaging effects to BM cells that result in the release of growth- and metastasis-promoting factors in vitro. Non-irradiated BM cells showed little or no effect on the growth of A2780 cells (upper panel). On the other hand, irradiated BM cells support the rapid and spontaneous growth of A2780 cell clusters (lower panel). b. Microscopic count of A2780 cell clusters found per well (24-well culture plate) after 48 hours of co-culture with irradiated and non-irradiated BM cells (left panel). Microscopic count of A2780 cell clusters after 48 hours of co-culture with CM samples harvested from BM, lung, and liver from irradiated and non-irradiated mice. (right panel). This co-culture experiment was repeated twice. In each experiment, six mice were employed per group (n = 3, no-irradiation control; n = 3, irradiation at 1000 cGy). Microscopic counts were made in triplicate from the wells, with significant results. Results are presented as means ± SD, with a statistical significance *p < 0.05, **p < 0.005, or ***p < 0.0005 relative to the control (A2780 cells cultured in RPMI 0.5% BSA medium).