In vivo efficacy assessment of the CDK4/6 inhibitor palbociclib and the PLK1 inhibitor volasertib in human chordoma xenografts

Abstract

Background: Management of advanced chordomas remains delicate considering their insensitivity to chemotherapy. Homozygous deletion of the regulatory gene CDKN2A has been described as the most frequent genetic alteration in chordomas and may be considered as a potential theranostic marker. Here, we evaluated the tumor efficacy of the CDK4/6 inhibitor palbociclib, as well as the PLK1 inhibitor volasertib, in three chordoma patient-derived xenograft (PDX) models to validate and identify novel therapeutic approaches.

Methods: From our chordoma xenograft panel, we selected three models, two of them harboring a homozygous deletion of CDKN2A/2B genes, and the last one a PBRM1 pathogenic variant (as control). For each model, we tested the palbociclib and volasertib drugs with pharmacodynamic studies together with RT-PCR and RNAseq analyses.

Results: For palbociclib, we observed a significant tumor response for one of two models harboring the deletion of CDKN2A/2B (p = 0.02), and no significant tumor response in the PBRM1-mutated PDX; for volasertib, we did not observe any response in the three tested models. RT-PCR and RNAseq analyses showed a correlation between cell cycle markers and responses to palbociclib; finally, RNAseq analyses showed a natural enrichment of the oxidative phosphorylation genes (OxPhos) in the palbociclib-resistant PDX (p = 0.02).

Conclusion: CDK4/6 inhibition appears as a promising strategy to manage advanced chordomas harboring a loss of CDKN2A/2B. However, further preclinical studies are strongly requested to confirm it and to understand acquired or de novo resistance to palbociclib, in the peculiar view of a targeting of the oxidative phosphorylation genes.

Keywords: CDK4/6 inhibitor; CDKN2A/2B deletion; PLK1 inhibitor; chordoma patient’s derived xenograft; palbocicib; volasertib.

Figure 1

Genome view profiles of the allele-specific copy number of the CD3 (A), CD7 (B), and CD39 (C) chordoma PDX models. The top graph log-odds-ratio represents the B-Allele Frequency (BAF), and the bottom graph the log depth ratio between tumor and a healthy witness. Both CD3 and CD7 chordoma PDX models harbored a homozygous deletion of CDKN2A/2B (red bar), whereas CD39 did not.

Figure 2

In vivo efficacy of palbociclib and volasertib, in the CD3, CD7, and CD39 chordoma PDXs. PDX tumor-bearing mice were randomized into each treatment group (n = 4–7 mice per group), and treated with palbociclib 75 mg/kg, 5 days per week (orange), or volasertib 10 mg/kg, 4 days per week (blue). Untreated control is shown in black. (A) Relative tumor volumes. Tumor growth was evaluated by plotting the mean of the relative tumor volume ± SD per group. (B) Probability of tumor progression. The time to reach RTV × 2 and RTV × 4 for each treated mouse has been calculated using Kaplan Meier curves and log-rank test.

Figure 3

Individual tumor responses in the three treated models. (A) Overall response rate (ORR) in all treated chordoma PDXs after palbociclib (n = 17) and volasertib (n = 16). A tumor was considered to be responding to treatment if ORR was below -0.5. (B, C) Probability of progression (RTV × 2 and RTV × 4) in all treated chordoma PDXs after palbociclib and volasertib treatments.

Figure 4

Gene expression modifications after treatments with palbociclib and volasertib (CD3 and CD7 models). Quantitative RT-PCR was performed to measure the expression of the target genes. Data are presented as level of gene expression compared to the standard level of Homo sapien (Hs) TBP. Genes expressions were compared between mice treated with the indicated agent and the control group. Error bars show the standard error of the mean. * and ** achieve statistical significance compared to control (p < 0.05 and p < 0.01, respectively) by Mann–Whitney U test.

Figure 5

Volcanoplots illustrating the top 1000 genes (A) and all genes related to the oxidative phosphorylation (OxPhos) pathway (B), in which we observed a significant gene expression variation in both untreated CD3 and CD7 PDX models (adjusted p-value).

Figure 6

The classic model for the regulation of the G1/S transition by cyclins and CDK. (A) In basal situation, RB protein phosphorylation performed by cyclin D-CDK4/6 kinase induces E2F releasing and facilitates the expression of E2F target genes, which are critical for initiation of DNA synthesis and entry into S-phase. CDK4/6 activity is repressed by the p16 protein. (B) Loss of p16/CDKN2A/2B activity leads to an overactivation of the CDK4/6 activity and therefore an uncontrolled cell proliferation. (C) The CDK4/6 inhibitor re-establishes the activity of the RB protein activity as an inhibitor of cell division.