Genetic Testing to Figure Out Drug Reactions

P Nicoletti et al. Gastroenterol 2023; 164: 454-466. Open Access! Identification of Reduced ERAP2 Expression and a Novel HLA Allele as Components of a Risk Score for Susceptibility to Liver Injury Due to Amoxicillin-Clavulanate

Key findings:

  • Transcriptome-wide association study revealed a significant association of AC-DILI risk with reduced liver expression of ERAP2 (P = 3.7 × 10–7), coding for an aminopeptidase involved in antigen presentation
  • “We also identified HLA-B∗15:18 as a novel AC-DILI risk factor in both discovery (OR, 4.19) and validation (OR, 7.78) cohorts”
  • A genetic risk score incorporating rs1363907, rs2476601, HLA-B∗15:18, HLA-A∗02:01, and HLA-DRB1∗15:01, was highly predictive of AC-DILI risk when cases were analyzed against both general population and non–AC-DILI control cohorts.
  • This genetic risk score does not apply to amoxicillin. “Clavulanate has long been considered the main component causing Amoxicillin-Clavulanate-DILI (LiverTox”
  • While the genetic risk score has high specificity, it is not highly sensitive. “Our multimarker model based on the 5 risk alleles predicted approximately 13% of the total risk for AC-DILI, which is approximately one-third of the total DILI risk that has previously been attributed to common genetic variants (approximately 40%)”

My take: In cases of suspected AC-DILI, identifying an abnormal genetic risk could help confirm the diagnosis. However, due to low sensitivity it is not likely to gain widespread clinical use.

Graphical Abstract:

Related blog posts:

Data on Drug-Induced Liver Injury

Two recent studies provide complementary information regarding the causes and consequences of Drug-Induced Liver Injury (DILI).:

  • Chalasani N, et al. Gastroenterol 2015; 148: 1340-52.
  • Goldberg DS, et al. Gastroenterol 2015; 148: 1353-61.

The first study looked at 899 patients with DILI in the DILI Network which is a consortium of several academic institutions funded by the US National Institutes of Health.  Antimicrobials were the most commonly implicated agents (408 cases); however, dietary/herbal supplements were another common cause (145 cases).  Top 10 individual agents:

  • Amoxicillin-clavulanate (Augmentin) (n=91)
  • Isoniazid (n=48)
  • Nitrofurantoin (n=42)
  • Sulfamethoxazole/trimethoprim (n=31)
  • Minocycline (n=28)
  • Cefazolin (n=20)
  • Azithromycin (n=19)
  • Ciprofloxacin (n=16)
  • Levofloxacin (n=13)
  • Diclofenac (n=12)

Key findings:

  • Overall, 10% of patients with DILI died or required liver transplantation.
  • 18% developed chronic injury pattern; this was more common in patients with a cholestatic liver injury.
  • Mortality was high in patients with DILI and concomitant severe skin reactions.  Causative agents of DILI with either Stevens-Johnson Syndrome or Toxic epidermal necrolysis included azithromycin (n=2), lamotrigine (n=3); and one case for each of the following: moxifloxacin, diclofenac, carbamazepine, nitrofurantoin, and possible cephalexin (patient rec’d lamotrigine as well)
  • Preexisting liver disease increased the likelihood of mortality (16% versus 5%)

The second article, a retrospective cohort study using data from >5 million covered individuals over a 7-year span from Kaiser Permanente Northern California, identified 62 inpatients categorized as having definite or possible acute liver failure (ALF).  In this cohort, 32 (52%) had DILI.  Leading agents of DILI-ALF:

  • Acetaminophen n=18
  • Herbal/dietary supplement n=6. Chinese herbals (n=2), pine needle tea, saw palmetto, one unspecified herbal.
  • Antimicrobials n=2

Bottomline: Antibiotics and herbal supplements, both of which are often used without apparent benefit, can lead to liver failure

Related blog posts:

Soapes Creek, Atlanta

Soapes Creek, Atlanta

More data on DILI

Using a population-based cohort, the authors of a recent study from Iceland performed a prospective study, collecting data on 96 individuals with drug-induced liver injury (DILI) from 2010-2011 (Gastroenterol 2013; 144: 1419-25 & editorial 1335-36).  This study benefitted from a centralized medical care system with about 250,000 adults.  Patients with acetaminophen toxicity were excluded.


  1. DILI occurred in 19 cases per 100,000 persons per year.  This is higher than previous DILI estimates in other populations and may be due to identifying more subclinical cases.
  2. DILI increased markedly with age from 9 per 100,000 among 15-29 year olds to 41 per 100,000 among those >70; however, this paralleled the increase use of medications.
  3. Eight drugs were implicated in more than 1 case with subsequent risk estimates:
  • Augmentin 1 per 2350 users
  • Azathioprine 1 per 133  (mostly anicteric cases)
  • Infliximab 1 per 148
  • Nitrofurantoin 1 per 1369
  • Isotretinoin 1 per 732
  • Atorvastatin 1 per 3693
  • Diclofenac 1 per 9148
  • Doxycycline 1 per 16,339

Some commonly implicated drugs did not show up on the list: fluoroquinolones, macrolide antibiotics, minocycline, valproic acid, and cancer chemotherapeutic agents (except one case due to imatinib).  The editorialist notes that these agents are not commonly used in Iceland.  Also, 16% of cases were due to herbals and dietary supplements.

Consequences of DILI: 27% developed jaundice, 12% required hospitalization, and 1 patient died.  The median time for liver tests to normalize in those that recovered was 64 days.

Related blog posts: