Looking at our data through genetic tinted glasses

The final plenary speaker for 2013 Kidney Week was Julie Johnson who discussed the treatment of hypertension and pharmacogenetics.

She started her discussion by discussing the human genome project. It was completed either ten or thirteen years ago, depending on whose history you believe. At the time of its completion it cost about one dollar per nucleotide and took over a decade. In 2011 it cost only $5-20k/genome and took a couple of weeks. Today it is less than $1,000 and takes less than a day. This geometric collapse in the price and time it takes to translate an individual’s genome is going to transform medicine. We are approaching the time when it will make sense to stop doing blood dots for the newborn screen and just sequence the child’s entire genome. That would diagnose genetic diseases and continue to pay dividends for the remainder of the child’s life.

Her personal area of research is the use of genetic information to target hypertension treatment. She started by pointing out how feeble our current treatment strategy is. We only get about 50% of patients to goal with our current empirical strategy. This low success rate leads to tremendous polypharmacy in the name of blood pressure control.

In order to research hypertension control she uses two types of data sets. The first are large hypertension trials with CV hard outcomes. She used INVEST as an example. The other are small short-term (4-10 weeks) trials where the outcome is blood pressure response.

In an example of the latter (small, short trials) she talked about the PEAR 1 trial. This was a trial of atenolol versus hydrochlorothiazide. Patients were randomized to one or the other drug and after a few weeks crossed over to the other drug. They had comprehensive genetic information and used that to to identify SNPs associated with drug responsiveness. Interestingly they found that the same SNP that was associated with HCTZ responsiveness, caused atenolol resistance. She didn’t feel that the SNPs she found were actually causative but just tightly linked on the genetic map.

Ultimately her team found 7 SNPs associated with HCTZ responsiveness. So patients can have zero to fourteen of these SNPs. The higher the dose of these SNPs the more HCTZ responsive the patient is. How useful would it be to know that the patient would be very sensitive or very resistant to HCTZ before the first dose?

Next, Johnson discussed the data they have from INVEST. 22,599 patients randomized to verapamil or atenolol. ACEi and diuretics were the add-on therapy for both arms of the study. The primary end point was a composite of death, stroke, or MI. There was no difference between arms in regard to the primary outcome. So anyone who does not like sub-group analysis of negative trials can stop reading. She combined the outcomes data with genetic data to generate a risk score based on three SNPs. Patients could be low or high risk. When she re-examined the primary intervention based on the genetic risk of the outcome she found the high-risk patients benefited from the beta-blocker, but the beta-blocker was not helpful for low-risk patients. On the other hand, verapamil was beneficial to low-risk patents and was harmful to high-risk patients. If we prescribe these drugs without knowing the patients’ genetic CV risk factor, we could just as well be harming them or helping them.

We are in the dark ages...and the lights are about to be turned on. #kidneywk13

— Joel M. Topf, MD FACP (@kidney_boy) November 10, 2013

Post written by Dr. Joel Topf, eAJKD Advisory Board member.

Check out all of the eAJKD coverage of ASN’s Kidney Week 2013!