Single-agent therapy with sorafenib or 5-FU is equally effective in human colorectal cancer xenograft--no benefit of combination therapy

Abstract

Background: We initiated this preclinical study in order to analyze the impact of sorafenib single treatment versus combination treatment in human colorectal cancer.

Methods: The effect of increasing sorafenib doses on proliferation, apoptosis, migration, and activation of signal cascades was analyzed in vitro. The effect of sorafenib single treatment versus 5-fluorouracil (5-FU) single treatment and combination therapy on in vivo proliferation and target cytokine receptor/ligand expression was analyzed in a human colon cancer xenograft mouse model using HT29 tumor cells.

Results: In vitro, SW480 and HT29 cell lines were sensitive to sorafenib, as compared to Caco2 and SW620 cell lines, independent of the mutation status of K-ras, Raf, PTEN, or PI3K. The effect on migration was marginal, but distinct differences in caspases activation were seen. Combination strategies were beneficial in some settings (sorafenib + 5-FU; irinotecan) and disadvantageous in others (sorafenib + oxaliplatin), depending on the chemotherapeutic drug and cell line chosen. Sensitive cell lines revealed a downregulation of AKT and had a weak expression level of GADD45β. In resistant cell lines, pp53 and GADD45β levels decreased upon sorafenib exposure. In vivo, the combination treatment of sorafenib and 5-FU was equally effective as the respective monotherapy concerning tumor proliferation. Interestingly, treatment with either sorafenib or 5-FU resulted in a significant decrease of VEGFR1 and PDGFRβ expression intensity.

Conclusions: In colorectal cancer, a sensitivity towards sorafenib exists, which seems similarly effective as a 5-FU monotherapy. A combination therapy, in contrast, does not show any additional effect.

Fig. 1

Effect of sorafenib monotherapy and combining sorafenib with standard chemotherapeutics 5-FU, oxalipaltin, or irinotecan. a While SW480 and HT29 cells were almost eradicated by high dose sorafenib treatment, low doses resulted in a significantly reduced growth as compared to placebo treatment. In contrast, Caco2 and SW620 cells were resistant to low-dose sorafenib while high-dose sorafenib stabilized tumor cell load of SW620 and Caco2. b 5-Fluorouracil. Combining 5-FU and sorafenib revealed additive effects (proliferation and apoptosis) in SW480 and HT 29 cells but not in SW620 or Caco2 cells. In contrast, the combination even increased proliferation of Caco2 cells. c Irinotecan. Combining 5-FU and irinotecan demonstrated clear additive effects in SW480, HT29 (proliferation and apoptosis), and SW620 (proliferation only) cells and even slightly increased apoptosis of Caco2 cells. However, no effect was seen in SW620 cells. d Oxaliplatin. Combining oxalipaltin and sorafenib revealed an additive effect (proliferation and apoptosis) in HT29 cells only but not in SW480, SW620, or Caco2 cells. In contrast, the combination increased proliferation of SW480, Caco2, and SW620 cells and inhibited apoptosis of SW480 and SW620 cells

Fig. 2

a Upon exposure with augmenting sorafenib doses, we observed an inhibition of the Ras–Raf pathway (pMEK) in Sw620 cell lines, only. In contrast, this pathway was activated in Caco2 cells. The AKT pathway was particularly altered in sensitive cell lines. SW480 and HT29 cells revealed only a hint–absent pAKT expression, but AKT expression was clearly suppressed upon exposure to increasing sorafenib doses. In contrast, the resistant cell line Caco2 did not show such AKT inhibitory behavior. An inhibition of pPI3K and pAKT was seen in Caco2 but not in SW620. b Analyses of the signaling pathways showed that sorafenib-sensitive cell lines reveal almost absent pAKT, absent–weak GADD45β, and medium–strong FoxO3a expression levels. In contrast, resistant cell lines showed a medium pAKT, very strong GADD45β, and weak–medium FoxO3a expression levels. GADD45β expression levels discriminated best between sensitive and resistant cell lines. When analyzing the impact of sorafenib, we observed that the sensitive cell line SW480 revealed a pp53 and a GADD45ß upregulation upon exposure to increasing sorafenib doses. In contrast, resistant cell lines showed initially high (Caco2, SW620) and, upon drug exposure, decreasing GADD45β (Caco2) and decreasing pp53 (Caco2, SW620) expression levels

Fig. 3

In vivo, only sorafenib monotherapy inhibited tumor growth significantly as compred to the control group. 5-FU treatment or the combination demonstrated only a nonsignificant inhibition. When all treatment groups were compared, no significant differences were observed

Fig. 4

a Analysis of Ki-67 proliferation index in three different treated animal groups (sorafenib monotherapy versus 5-FU monotherapy versus combination therapy) show a significant reduction of proliferation in all treatment groups but no superior therapeutic effect of the combination therapy group. b Interaction plot for Ki-67 depicts no additive effect of combination therapy with 5-FU and sorafenib. Analysis of the expression intensity of PDGFA, PDGFRβ, and VEGFR1 in the control group and under treatment with sorafenib monotherapy, 5-FU monotherapy, and combination therapy. PDGFRβ and VEGFR1 expression intensities were significantly reduced by 5-FU and sorafenib, respectively

Fig. 5

a Analysis of the expression intensity of PDGFA, PDGFRβ, and VEGFR1 in the control group and under treatment with sorafenib monotherapy, 5-FU monotherapy, and combination therapy. PDGFRβ and VEGFR1 expression intensities were significantly reduced by 5-FU and sorafenib, respectively. b Immunostaining was evaluated by three authors independently. The immunohistochemical staining was analyzed according to a scoring method as previously validated and described by Laverdiere et al. The tumors were classified into four groups based on the homogeneous staining intensity: 0, no expression; 1, weak expression; 2, moderate expression; and 3, strong expression. In the case of heterogeneous staining within the same sample, the respective higher score was chosen if more than 50 % of cells revealed a higher staining intensity. If evaluations did not agree, specimens were re-evaluated and reclassified according to the assessment given most frequently by the observers