Synergistic drug combinations and machine learning for drug repurposing in chordoma

Area of Science:

• Oncology
• Pharmacology
• Computational Biology

Background:

• Chordoma is a rare, aggressive cancer with poor survival and no approved drugs.
• Current treatments include surgery and radiation, highlighting the need for novel therapeutic strategies.
• Drug repurposing accelerates the development of new cancer medicines by using compounds already tested for safety.

Purpose of the Study:

• To identify potential drug candidates for chordoma through a drug repurposing strategy.
• To validate machine learning models for predicting chordoma inhibition.
• To test combinations of approved kinase inhibitors for synergistic effects in chordoma.

Main Methods:

• Generated and validated Bayesian machine learning models using published chordoma cell line inhibition data.
• Screened compounds from the NIH NCATS industry-provided assets using the developed models.
• In vitro testing of identified compounds and drug combinations, including AZD2014 (mTOR inhibitor), afatinib (EGFR inhibitor), and palbociclib (CDK4/6 inhibitor).

Main Results:

• The mTOR inhibitor AZD2014 showed potent inhibition against chordoma cell lines (IC50 values of 0.35 µM and 0.61 µM).
• Combinations of approved drugs demonstrated synergistic effects: afatinib and palbociclib (CI50 = 0.43), and AZD2014 and afatinib (CI50 = 0.41).
• The mTOR signaling pathway is implicated as a promising target for chordoma therapy.

Conclusions:

• Drug repurposing and computational approaches can effectively identify novel therapeutic candidates for rare cancers like chordoma.
• AZD2014 and synergistic combinations of approved kinase inhibitors represent promising leads for chordoma treatment.
• The in vitro and in silico methodology used can be applied to accelerate drug discovery for other rare malignancies.