15 May 2012. A common variant near the gene encoding the melanocortin 4 receptor (MC4R) is associated with extreme weight gain in people taking antipsychotics, according to a study published online May 7 in the Archives of General Psychiatry. Led by Anil Malhotra of Zucker Hillside Hospital in Glen Oaks, New York, the report includes the first genomewide association study (GWAS) of a cohort monitored as they took second-generation antipsychotics for the first time. In this group, and in three independent cohorts, people carrying two copies of the low-frequency allele gained about twice as much weight as those with one or no copies did.
The findings seem to deliver on the promise of pharmacogenomics in predicting drug response, in this case a side effect. Second-generation antipsychotics are notoriously associated with weight gain, which compromises health and decreases motivation to continue taking the medication. Last year, a GWAS of metabolic side effects to antipsychotics in chronically treated adults with schizophrenia identified several genes of interest, but not MC4R (Adkins et al., 2011).
To narrow in on metabolic effects strictly due to antipsychotic drugs, the new study focuses on people taking antipsychotic drugs for the first time for a range of disorders not limited to schizophrenia. Malhotra has previously documented extreme weight gain in a subset of pediatric patients upon their first exposure to second-generation antipsychotics (Correll et al., 2009): after just 12 weeks on the drugs, one-quarter of the patients gained a hefty 15 to 35 lbs. This prompted Malhotra and colleagues to design a GWAS that might genetically identify such people, which could eventually be used to inform medication decisions.
All hits lead to MC4R
First author Malhotra started with DNA samples from 139 pediatric patients from the previous study, which were then genotyped on a chip with about one million SNPs. Not only were the subjects antipsychotic-naïve at the start of the study, but once they started medication, their compliance was confirmed with blood tests over 12 weeks. People taking olanzapine were excluded because it conferred a different weight gain profile.
The GWAS delivered 20 single nucleotide polymorphism (SNP) hits in the same region of chromosome 8, though none reached genomewide significance (all P <10-5). However, the region has already been implicated in obesity and body mass index in the general population (Loos et al., 2008). About 190 kb downstream from these GWAS signals lies the MC4R gene, which encodes a protein involved in feeding behavior, and when disrupted, causes extreme obesity in humans and mice (Farooqi et al., 2003; Huszar et al., 1997). The team found that, for each of these 20 SNPs, carriers homozygous for the minor allele gained significantly more weight than did heterozygotes or subjects homozygous for the more common allele, though these latter two groups also gained weight.
To validate these results, the researchers genotyped three of these SNPs in three additional, independent cohorts (see SRF related news story). Each cohort consisted of people with schizophrenia taking second-generation antipsychotics for the first time, and for at least six weeks, their weight gain and adherence to medication was monitored. Of these, one SNP (rs489693) produced a significant effect for all three groups, again with two copies of the minor allele associated with weight gain about twice that of people with one or no copies. A meta-analysis of all four groups, with a combined sample size of 344, yielded a genomewide significance for this SNP (P = 5.59 x 10-12). Baseline weight and body mass index did not correlate with the amount of weight gained.
The researchers also reported that, in the discovery cohort, this same SNP modulated other metabolic measures such as triglycerides, insulin, and total fat mass in the same recessive manner.
Out in the real world
Though some might argue that the stringent approach taken in this study is not informative for the real world, where people with schizophrenia frequently don't take their medications, it may have given a dataset clean enough—uncontaminated by people who didn't gain weight just because they didn't take their medications—to detect a consistent signal near MC4R. That this signal emerged despite the fact that the sample size was small by GWAS standards may reflect the use of a tightly controlled drug exposure and a reliably measured phenotype. The researchers suggest that similar “environmental challenge” approaches in which an environmental factor is introduced while an outcome is measured may help detect susceptibility alleles in other realms.
Applying the results to the real world, the researchers suggest that the subset of patients homozygous for the minor allele at this SNP could also take MC4R agonists to help mitigate the metabolic downsides of second-generation antipsychotics. In the meantime, researchers may find fertile ground in the biology of how these antipsychotics interact with MC4R, which could further inform treatment strategies.—Michele Solis.
Malhotra AK, Correll CU, Chowdhury NI, Müller DJ, Gregersen PK, Lee AT, Tiwari AK, Kane JM, Fleischhacker WW, Kahn RS, Ophoff RA, Lieberman JA, Meltzer HY, Lencz T, Kennedy JL. Association Between Common Variants Near the Melanocortin 4 Receptor Gene and Severe Antipsychotic Drug-Induced Weight Gain. Arch Gen Psychiatry. 2012 May 7. Abstract