Thiopurine Pathway updated with new candidate genes

Thiopurines and TPMT have long been a cornerstone of PGx research. As such the Thiopurine Pathway was one of the earliest pathways in PharmGKB to be published in Pharmacogenetics and Genomics [PMID:19952870]. In collaboration with Jun Yang, Takaya Moriyama and Rina Nishii from St Judes, we have substantially updated the Thiopurine Pathway to include new candidate genes in particular NUDT15. NUDT15 will also be featured in a new VIP summary next month. 

The new Thiopurine Pathway figure focuses more on the drug metabolites (which are clickable and linked to Pubchem structures and publications) and candidates for their formation. We have also updated the text to include the kind of summary and literature review PharmGKB pathways are known for, and have a more detailed underlying set of components to allow for data export/analysis. 

Introducing the new PharmGKB Cancer PGx Portal

PharmGKB has collected a number of resources for Cancer PGx into one easy location. There are tables with direct links to genes important for cancer drug response both for PD and PK, to cancer drug pathways, particular cancers that have PGx data, types of toxicities common to cancer drugs, and external resources.

Eight new VIP gene pages give a short text based summary of important genes for cancer drug response. These are for the genes ALK, ABL1, BCR, BRAF, ERBB2 (HER2), KIT, KRAS and NRAS. Anyone with expertise in the genes who wishes to develop these with us for publication in PG&G, please contact feedback.

There is a shortlist of drug labels for cancer drugs with biomarker PGx.

We currently have 34 anti-cancer agent drug pathways with 8 new pathways in development. The portal gives shortcuts to a selection.

PharmGKB currently uses a flat ontology for diseases, which means that the Neoplasms disease page does not link to the many different cancers we have data for. The cancer portal has direct links to the cancers for which there is the most PGx information in the knowledgebase, such as pediatric ALL, CML, colorectal, breast, renal and non-small cell lung cancers. The portal also has links to the common types of toxicities with PGx data.

Finally there is a collection of external links that are useful for Cancer PGx.

Cisplatin FDA label changes

In 2013 we blogged about the controversy with the FDA labeling for cisplatin and its warning for TPMT variants and increased risk of ototoxicity in children. The 2012 label stated:

“Certain genetic variants in the thiopurine S-methyltransferase gene (e.g., TPMT*3B and TPMT*3C) are associated with an increased risk of ototoxicity in children administered conventional doses of cisplatin...Children who do not have one of these TPMT gene variants remain at risk for ototoxicity. All pediatric patients receiving cisplatin should have audiometric testing at baseline, prior to each subsequent dose, of drug and for several years post therapy.”

The label change appeared to be based on a single study from 2009.  Several years have passed and there is still very limited data in the public domain about TPMT and risk for cisplatin-induced ototoxicity: four papers comprising seven studies and a meta-analysis [PMID:19898482] [PMID:23820299] [PMID:23588304] [PMID:25551397]. The meta-analysis found no association with any of the TPMT variants and that the studies were significantly heterogeneous in terms of cancer types, ethnicities, age and co-treatments, all of which influence risk for ototoxicity.

The FDA has requested changes to the cisplatin label removing the whole Pharmacogenomics section, to now state :

“Genetic factors (e.g. variants in the thiopurine S-methyltransferase [TPMT] gene) may contribute to cisplatin-induced ototoxicity; although this association has not been consistent across populations and study designs.” 


This better reflects the current uncertainty about this gene-drug relationship and is more in line with the level of evidence we have assessed in our clinical annotations for the TPMT variants and cisplatin. 

Combining large PGx datasets from cancer cell lines

Testing cancer cell lines in vitro for drug sensitivity is a cornerstone of preclinical drug development. Large publically available datasets can be found at The Genomics of Drug Sensitivity in Cancer Project (GDSE) and The Cancer Cell Line Encyclopedia (CCLE).

Studies attempting to combine large public datasets and analyzing for correlation questioned the reliability of the data due to limited concordance, reported in [PMID: 24284626], discussed in [PMID:24284624] and a confirmation study here.

A new report in Nature describes different methods to analyze the data from CCLE and GDSE and concludes that “data from either study yields similar predictors of drug response” [PMID:26570998].

These papers demonstrate the continuing difficulty trying to compare across large datasets. Such problems include comparing different experimental protocols and measurements for drug sensitivity across studies, trouble matching the drug and cell line names to ensure like comparison, discordance in the genotyping data, and drugs that had few examples of cell lines that were drug sensitive.  As always, attention to detail in the documentation and description of the experiments can help mitigate some of these difficulties. While development of standard testing protocols and data curation and reporting frameworks may lead to better validation of drug response predictors going forward there will always be the need for methods to filter the noise that is inevitable in large datasets.