|dc.description||MATERIALS AND METHODS
A total of 870 DILI cases were evaluated, which were induced by a wide range of drugs and dietary supplements. In the present study, 701 cases were included, while 169 were excluded due to not meeting the study inclusion criteria. This cohort was used to evaluate host factors and pharmacological drug properties in search for those associated with the development of different features in DILI (type of liver injury, delayed onset and severity) using the same methodology for all three parts. Hence, only cases induced by conventional drugs were considered for this study. Host factor information was retrieved from the Spanish DILI registry database. Drug property information was retrieved from the Liver Toxicity Knowledge Base (LTKB), a database developed by the National Center for Toxicological Research. Univariate analyses and nominal logistic regression were performed using the JMP (Cary, NC, USA) and R software to compare causative drug properties and host factors.
Type of liver injury
Seven parameters were identified as predictors for different types of liver injury in DILI. The factors selected by the logistic regression as enhancers of the hepatocellular injury were as follows: age below 60 years (OR 2.06, confidence interval to 95 % (95%CI): 1.28-3.34, p=0.0029), previous allergic reactions to drugs (OR 2.69 [95% IC: 1.01-8.24], p=0.0493), culprit drug with enterohepatic circulation (OR 4.94 [95% IC: 2.34-11.45], p=0.0001), drugs with high hepatic metabolism, (OR 1.70 [95% IC: 1.02-2.84], p=0.0417), sp3/sp2 hybridization ratio (<50%). Protective parameters were: intake of concomitant medication for treatment of cardiac diseases (OR 0.47 [95% IC: 0.29-0.77], p=0.0030) and increased number of heterorings in the drug molecule (OR 0.68 [95% IC: 0.52-0.88], p=0.0270). The ROC curve showed an AUC of 0.75863. A value between 0.75-0.90 indicates that the model has a good discrimination capacity.
In addition to the individual host factors and drug properties identified as liver injury predictors, several interactions between these factors also seem to play a role in the modulation of type of liver injury. Drugs with enterohepatic circulation increased the risk of hepatocellular type of injury in DILI patients with previous allergic reactions to drugs. In contrast, drugs with high hepatic metabolism decreased the risk of hepatocellular damage in patients with previous allergic drug reactions. Elevated number of heterorings was found to play a protective role in cases with previous allergic drug reaction. In addition, hybridization ratio was found to have a contradictory effect and increased the risk of hepatocellular injury in patients with previous allergic reactions to drugs. Furthermore, age seems to play an important role in modulating type of liver injury. Low hepatic metabolism (<50%) or lack of enterohepatic circulation increased the frequency of cholestatic cases to approximately 50% in older patients.
Three variables were found to increase the likelihood of delayed onset: absence of associated diseases (OR 1.91 [95% IC: 1.08-3.39], p=0.0245), culprit drug daily dose ≥200mg (OR 2.89 [95% CI: 1.25-7.58], p=0.0123) and hepatic metabolism <50% (OR 23.98 [95% CI: 10.84-63.68], p=0.0001).The calculated R2 for this model was 0.3034 and the area under the ROC curve 0.8459. The interaction study resulted in four pairs of potential interactions that seem to have significant impacts on delayed onset. Three host factors, age (cut-off 60 years), gender, and pre-existing cardiac diseases, were identified as significant effect modifiers for drugs with mitochondrial liability. Pre-existing cardiac diseases were also identified as a significant effect modifier for drugs with high Pauling electronegativity.
Only concomitant use of anti-infectious drugs was found as a predictor for high severity. Consumption of anti-infective agents prior to or at the time of the episode doubled the risk of progressing to a highly severe outcome (OR 2.00, IC: 1.35-3.47, p=0.008). Due to the fact that a single predictor (concomitant use of anti-infectives) was found to explain the risk of severity, no further analyses of host factor – drug property interactions were performed.
CONCLUSION 1: TYPES OF LIVER INJURY
a. Within the host factors, age is the determining factor of the phenotypic expression of the damage. Hepatocellular damage predominates in young patients, whereas patients aged ≥ 60 years present a higher frequency of cholestatic damage. In addition, the history of cardiovascular disease would condition a greater susceptibility to the development of cholestatic damage.
b. The main drug properties that modulate the expression of toxic liver damage are liposolubility, degree of binding protein, extent of hepatic metabolism and presence of enterohepatic circulation. Hepatocellular damage would be determined by drugs that have higher lipophilicity, high plasma protein binding and significant hepatic metabolism. Alternatively, cholestatic damage would be related to drugs more soluble in water, with poor protein binding and poor hepatic metabolism.
c. Other physicochemical characteristics identified as potential modulators of the damage are the presence of rings with heteroatoms in the chemical structure, the presence of sulfur atoms and / or the hybridization rate of orbitals.
d. The interaction between host factors and drug properties suggests that advanced age and consumption of drugs with poor hepatic metabolism increase the risk of cholestatic damage in susceptible individuals. Contrary to expectations, in young patients the risk of developing hepatocellular damage does not change depending on the rate of hepatic metabolism.
CONCLUSION 2: DELAYED INITIATION OF SYMPTOMS
a) Drugs with potential to induce delayed onset have shorter half-life, lower hepatic metabolism and have a higher proportion excreted in parent form than drugs without delayed onset potential. These properties would be related to higher solubility, lower lipophilicity and lower plasma protein binding.
b) Three independent factors appear to influence the likelihood of delayed symptom onset: hepatic metabolism (drug), daily dose (host) and pre-existing diseases (host).
c) Results derived from the interactions between host factors and pharmacological drug properties suggest that the interaction between mitochondrial liability and age, sex or underlying cardiac diseases modulates the risk of hepatotoxicity with delayed onset of symptoms. Pauling electronegativity combined with underlying cardiac diseases similarly affects the probability of delayed onset.
CONCLUSION 3: SEVERITY OF LIVER DAMAGE
a) Drugs with high hepatic metabolism and a higher rate of metabolized drug excretion seem to contribute to the increase in severity. This observation could be related to the higher frequency of hepatocellular cases, which are associated with a greater severity of the damage.
b) The presence of infectious diseases and / or the consumption of anti-infective comedication increase the risk of developing a more severe hepatic toxicity episode.
Finally, the identification of host factor-drug property interactions may provide insight into the underlying mechanisms responsible for the heterogeneity of clinical presentations in DILI. Furthermore, the generated hypotheses in the current study should be further explored in independent DILI cohorts.||en_US
Drug-induced liver injury is an important concern in public health. In addition, it is one of the most common adverse events that cause lack of drug approval, drug withdrawals, and post-marketing regulatory decisions despite its low frequency in the general population. Nowadays, this adverse reaction is an important challenge during clinical phases in drug development. Hepatotoxicity is considered a multifactorial pathology dependent on host, drug and environmental aspects. This aspect combined with the wide range of clinical presentation associated with DILI, makes it difficult to characterize specific factors that play a role in DILI development. For this reason, the current study approach takes into consideration that both drug properties and host factors could modulate different features in DILI. Identifying specific parameters within these groups that could potentially have a modulatory role in DILI may help to understand the underlying mechanism(s) in DILI. In addition, the detection of such parameters could help prevent a fatal outcome for patients with this pathology. Based on previous observations, drug properties such as hepatic metabolism and lipophilicity may play an important role in delayed onset, liver injury pattern and DILI severity. Furthermore, these properties are likely to be modulated by host factors.||en_US