Skip to main content

PanDrugs System Aims to Advance Precision Cancer Care by Matching Treatments to Mutations


CHICAGO (GenomeWeb) – A Spanish team of biomedical and cancer informatics specialists has developed a computational methodology to help clinicians and researchers prioritize potential cancer treatments based on personal genomics.

This platform, called PanDrugs, identifies "druggable" genes in patients with prostate, breast, and colorectal cancer who have had their tumors sequenced. PanDrugs then recommends therapies according to how effective each drug has proven to be against each known cancer gene. By assigning scores to drugs and genes, the methodology also has shown promise for treating patients without known druggable mutations.

The researchers, primarily from the Spanish National Cancer Research Centre, described their work in an article published today in Genome Medicine.

PanDrugs is "a new computational resource to propose drug therapies from genome-wide experimental results, including variant and gene lists," they wrote. "PanDrugs expands cancer therapeutic options by taking into account multiple genomic events potentially responsive to a treatment, the pathway context, and the pharmacological evidence reported in large-scale experiments."

The system draws on a database — dubbed PanDrugsdb — that integrates data from 24 different genomic, drug, and pharmacogenomic resources.

PanDrugs is focused exclusively on cancer. "The idea is, we analyze the function and impact of these genes that are mutated from this patient, and then we match this genetic profile with the database and find which of the therapies or treatments should be the most appropriate," said corresponding author Fátima Al-Shahrour, head of bioinformatics at the National Cancer Research Centre in Madrid.

Meant to be a reference tool for precision cancer care, the database grows as new genomic and pharmacogenomic information becomes available. PanDrugsdb currently lists 56,297 drug-target associations from a total of 4,804 genes and 9,092 pharmaceutical compounds, according to the PanDrugs website.

System architects standardized drug names in the database, then annotated known gene-drug relationships with information about genomic alterations associated with drug responses.

"The PanDrugs method has been implemented to address the interpretation gap between raw genomic data and clinical usefulness," the researchers explained in the Genome Medicine article.

PanDrugs breaks down druggable genes into three categories: direct targets; biomarkers; and "pathway members."

"When you apply genomics or targeted panels of genes in the clinical setting, there some genes that are very well known as drivers, and … they are proposed as targeted therapies," Al-Shahrour said. Certain frequently mutated genes are particularly important here, she said. "When you have this kind of gene mutation, you can have a specific treatment that is going to inhibit the activity" of a cancer.

For example, the journal article noted, the BRAF gene is a direct target for vemurafenib.

The biomarkers category includes genes associated with a drug response, but the target is not a specific protein. For example, BRCA-mutated cancers that respond to PARP inhibitors fall into this category, the paper noted.

Pathway members are downstream druggable genes that take advantage of therapeutic pathways underlying an individual's genome. "Interestingly, this paradigm unlocks alternative therapeutic ways for untargetable genes," the researchers wrote.

In concert with this categorization, PanDrugs calculates two values — the Gene Score (GScore) and the Drug Score (DScore) — based on the PanDrugsdb database and on patient-specific variant lists to make therapeutic recommendations. "PanDrugs expands the anticancer therapeutic arsenal suggesting drugs to target genes located downstream to the altered gene(s)," the article said.

While primarily designed for clinical care, PanDrugs also can be used for research with laboratory animals or human subjects, Al-Shahrour said. Scores help show the relevance of mutations.

"If this mutation is important in terms of the biology — that is the GScore — we consider that is the gene probably that we should inhibit," Al-Shahrour explained. "So then, we identify which drug should be the best one to inhibit this gene, and then we calculate the Drug Score."

Via the DScore, the system prioritizes drugs that have gained regulatory approval over those still in trial or experimental phases, according to Al-Shahrour. "If we are going to treat a patient, our priority is to find something that, of course, is approved or in clinical trial," she explained.

"But the reality is, when you analyze a patient, most of the mutations that you even define, you don't know exactly what they are doing. You don't know the function or the implication of these mutations," Al-Shahrour said. PanDrugs is intended to shed some light on these mysteries.

As with any clinical decision support system, PanDrugs recommendations do come with a disclaimer. "Nothing is validated. It's more or less to suggest or to help with the mission to propose which is going to be the best therapy," Al-Shahrour said.

"PanDrugs is the first method to systematically infer novel targeted treatments following a rational framework supported by multi-gene markers, molecular pathway context, and pharmacological evidence," the paper concludes.

"Our results show that in silico prescription approaches focused uniquely on known cancer genes should be complemented by incorporating drug information associated to genomic alterations located in non-cancer genes. Our approach extends the treatment opportunities of cancer patients by enriching the therapeutic arsenal against tumors and opens new avenues for personalized medicine."

Al-Shahrour discussed PanDrugs at the 2018 American Association for Cancer Research meeting in April, but with the publication of this new paper, the technology and database now are public. Since the project was funded in part by a European Commission Marie Curie Career Integration Grant, and because the National Cancer Research Centre collaborated with other institutions in Spain, PanDrugs has been made available to the open-source community, and Al-Shahrour is pushing the research possibilities of the technology.

"We need more clinical trials with genomic profiles to test these specific tools," she said. Indeed, a clinical trial sponsored by the European Research Council is underway to test the methodology in patients with pancreatic cancer.