Skip to main page content
U.S. flag
See More

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Clinical Trial
. 2024 Aug 1;4(8):2058-2074.
doi: 10.1158/2767-9764.CRC-24-0113.

mTORC1-Driven Protein Translation Correlates with Clinical Benefit of Capivasertib within a Genetically Preselected Cohort of PIK3CA-Altered Tumors

Constance A Sobsey  1   2 , Bjoern C Froehlich  3   4 , Georgia Mitsa  1   2 , Sahar Ibrahim  1   2 , Robert Popp  5 , Rene P Zahedi  1   6   7   8   9 , Elza C de Bruin  10 , Christoph H Borchers  1   2   11   12 , Gerald Batist  13   14
Affiliations
Clinical Trial

mTORC1-Driven Protein Translation Correlates with Clinical Benefit of Capivasertib within a Genetically Preselected Cohort of PIK3CA-Altered Tumors

Constance A Sobsey et al. Cancer Res Commun. .

Abstract

Capivasertib is a potent selective inhibitor of AKT. It was recently FDA approved in combination with fulvestrant to treat HR+, HER2-negative breast cancers with certain genetic alteration(s) activating the PI3K pathway. In phase I trials, heavily pretreated patients with tumors selected for activating PI3K pathway mutations treated with capivasertib monotherapy demonstrated objective response rates of <30%. We investigated the proteomic profile associated with capivasertib response in genetically preselected patients and cancer cell lines. We analyzed samples from 16 PIK3CA-mutated patient tumors collected prior to capivasertib monotherapy in the phase I trial. PI3K pathway proteins were precisely quantified with immuno-Matrix-Assisted Laser Desorption/Ionization-mass spectrometry (iMALDI-MS). Global proteomic profiles were also obtained. Patients were classified according to response to capivasertib monotherapy: "clinical benefit (CB)" (≥12 weeks without progression, n = 7) or "no clinical benefit (NCB)" (progression in <12 weeks, n = 9). Proteins that differed between the patient groups were subsequently quantified in AKT1- or PIK3CA-altered breast cancer cell lines with varying capivasertib sensitivity. The measured concentrations of AKT1 and AKT2 varied among the PIK3CA-mutated tumors but did not differ between the CB and NCB groups. However, analysis of the global proteome data showed that translational activity was higher in tumors of the NCB vs. CB group. When reproducibly quantified by validated LC-MRM-MS assays, the same proteins of interest similarly distinguished between capivasertib-sensitive versus -resistant cell lines. The results provide further evidence that increased mTORC1-driven translation functions as a mechanism of resistance to capivasertib monotherapy. Protein concentrations may offer additional insights for patient selection for capivasertib, even among genetically preselected patients.

Significance: Capivasertib's first-in-class FDA approval demonstrates its promise, yet there remains an opportunity to optimize its use. Our results provide new evidence that proteomics can stratify genetically preselected patients on clinical benefit. Characterization of the same profile in cell lines furnishes additional validation. Among PIK3CA-altered tumors, increased mTORC1-driven translation appears to confer intrinsic resistance. Assessing mTORC1 activation could therefore prove a useful complement to the existing genetic selection strategy for capivasertib.

PubMed Disclaimer

Conflict of interest statement

R.P. Zahedi reports personal fees from MRM Proteomics Inc. outside the submitted work. E.C. de Bruin reports AstraZeneca employee and holds AstraZeneca shares. C.H. Borchers reports grants from Genome Canada and Genome Quebec, and other support from AstraZeneca during the conduct of the study; C.H. Borchers is the C.S.O. of MRM Proteomics, Inc. and the V.P. of Proteomics at Molecular You outside the submitted work. No other disclosures were reported.

Figures

Figure 1
Figure 1
Statistical analysis of targeted and global proteome comparing protein expression in analyzed FFPE tumor tissues from the CB (green) and NCB (red) groups. A, Schematic of experimental design for phase I of the study. B, Results of targeted quantitation by iMALDI-MS. Boxplots of protein concentrations in the CB (green) vs. NCB (red) groups. Each point represents one patient tumor, averaged for multiple slides. P-values are given for a two-tailed t test. C, Heatmap of 50 selected proteins with highest differential expression between groups showing their normalized LFQ abundances across both analyzed batches. n = 23. Arbitrary identifiers are included for the purpose of identifying replicates from the same tumor. D, Volcano plot showing the fold change (log2FC) of all protein expression features vs. P-value (−log10P). E, PCA. Proximity of sample replicates originating from the same patients is indicated with a (dashed yellow line). F, Features ranked by VIP based on their contribution to the discrimination between the CB and NCB groups in the PLS-DA. The arrows on the right of the VIP Feature list show relative expression of that protein in the NCB group vs. the CB group.
Figure 2
Figure 2
Mapping of global proteomics results from CB vs. NCB groups to protein networks and pathways. A, Network analysis via String-DB plot of high-confidence protein–protein interactions among proteins that are significantly different between CB and NCB group. Edges represent lines of evidence of protein–protein interactions. Line colors indicate different evidence types: known interactions from curated databases (teal) or published experiments (pink), predicted interactions based on gene location (green), fusion (red), or co-occurrence (blue), or observed co-expression (black), protein homology (purple), or textmining relationships (lime). Major clusters are labeled according to shared features within the cluster. Nodes are overlaid from the Cytoscape visualization, with fold-change shown by color. Node size increases as the P-value decreases. B, Top 20 canonical pathways that are significantly differentially activated between clinical benefit (CB) and no clinical benefit (NCB) groups, based on assessment with Fisher’s exact test in Qiagen IPA. The height of the bar corresponds to the confidence of an association, with a threshold of P < 0.01. IPA’s Z-score indicates the direction of regulation and extreme z-scores are depicted with increased color intensity. Orange bars represent increased activation in the NCB group whereas blue bars represent upregulated activity in the CB group. White bars indicate pathways with fewer than four mapped proteins or z-scores close to zero, indicating that the direction of regulation of individual pathway members does not strongly match a prespecified pattern. Gray bars indicate pathways for which no prediction can be made due to available evidence in the database.
Figure 3
Figure 3
Statistical analysis of targeted protein quantitation in capivasertib-sensitive (green) and capivasertib-resistant (red) HR+PIK3CA-altered breast cancer cell lines. A, Schematic of experimental design for phase II of the study. B, Heatmap of protein concentrations quantified by MRM-MS with hierarchical clustering. C, Cell line sensitivity to capivasertib, expressed as IC50 (D) Protein concentrations in capivasertib-sensitive vs. capivasertib-resistant cells (E) PCA (left) and PLS-DA plot (right). F, Top 22 features ranked by VIP scores, presented with the direction of change in capivasertib-resistant vs. capivasertib cell lines as compared to NCB vs. CB patient tumors.
Figure 4
Figure 4
Schematic depicting proteins and pathways differently regulated in capivasertib-resistant cancers, as observed in patient tumors and breast cancer cell lines. Proteins with expression changes confirmed (FC > 1.3; P < 0.1) by MRM-MS data in cell lines are shown in red (increased), dark gray (no change), and green (decreased). Proteins with expression data from LFQ in patient tumor samples are shown in dark orange (increased) and dark blue (decreased). Light orange (activated) and light blue (inhibited) indicate proteins with predicted activity changes based on pathway analysis of the LFQ data. Proteins with no expression data are shown in light gray. (Figure Created with BioRender.com.)
See this image and copyright information in PMC

References

    1. Millis SZ, Ikeda S, Reddy S, Gatalica Z, Kurzrock R. Landscape of phosphatidylinositol-3-kinase pathway alterations across 19 784 diverse solid tumors. JAMA Oncol 2016;2:1565–73. - PubMed
    1. Engelman JA. Targeting PI3K signalling in cancer: opportunities, challenges and limitations. Nat Rev Cancer 2009;9:550–62. - PubMed
    1. De Luca A, Maiello MR, D’Alessio A, Pergameno M, Normanno N. The RAS/RAF/MEK/ERK and the PI3K/AKT signalling pathways: role in cancer pathogenesis and implications for therapeutic approaches. Expert Opin Ther Targets 2012;16(Suppl 2):S17–27. - PubMed
    1. Nitulescu GM, Margina D, Juzenas P, Peng Q, Olaru OT, Saloustros E, et al. . Akt inhibitors in cancer treatment: the long journey from drug discovery to clinical use (Review). Int J Oncol 2016;48:869–85. - PMC - PubMed
    1. Turner NC, Oliveira M, Howell SJ, Dalenc F, Cortes J, Gomez Moreno HL, et al. . Capivasertib in hormone receptor–positive advanced breast cancer. N Engl J Med 2023;388:2058–70. - PMC - PubMed

Publication types

MeSH terms