Despite the perception that GA has low toxicity to humans and the environment [13], GA poisoning remains a significant health issue in our country due to the delayed onset of symptoms. Many patients with GA poisoning present to the emergency department without any apparent symptoms but later develop complications during diagnosis and treatment. Unfortunately, there are currently no reliable early indicators that can accurately predict the prognosis of patients with GA poisoning.
In this study, we provide a summary of the clinical features of acute oral GA poisoning. Nausea and vomiting were the most common symptoms experienced by 80% of patients within 6 hours after ingesting the herbicide, additionally, herbicides can also cause gastrointestinal irritation. The majority of patients recovered from these symptoms within 1–2 days after hospitalization. However, neurological complications, such as coma, dizziness, lethargy, and seizure, occurred in 27 patients (40.3%), and were not apparent during the first few hours of hospitalization. The incidence of neurotoxicity was lower than the results of the research conducted by Lee DK [14]. In their research, the incidence of neurotoxicity was 64.4%, with only the latent periods of seizure was studied. We first report that the different latent periods of different symptoms of neurotoxicity. The median time for consciousness and convulsion or seizure was 33.0 (26.0, 40.0) hours and 45.0 (37.5, 48.5) hours after ingesting GA, respectively. Inhibition of glutamine synthetase by glufosinate ammonium may lead to elevated blood ammonia levels that are related to neurotoxicity symptoms [15]. Glutamine synthetase is a crucial enzyme in the human body that plays a critical role in brain metabolism and the normal function of various organs[16]. Hyperammonemia caused by herbicides was difficult to decrease through lactulose, which is different from the elevated serum ammonia levels associated with hepatic encephalopathy[17]. It is important to monitor respiratory function within 24 hours of taking the herbicide, as 11 patients (16.4%) suffered from respiratory failure during hospitalization, with 80% of these cases occurring within 12 hours after admission. The mechanism of respiratory failure is not well understood, but it should be closely monitored in the early stages of hospitalization. The overall mortality rate was 13.4% (9 patients), with hyperammonemia and hemodynamic toxicity being the main mechanisms of GA poisoning in humans [18–19]. The mortality of GA poisoning patients was rarely reported in the literature. Neurotoxicity symptoms and central apnea requiring mechanical ventilation support were thought to be caused by hyperammonemia and hemodynamic toxicity [20]. Although previous study [21] have found significant correlations between ammonia levels in the serum and CSF and neurological complications, CSF ammonia detection was not available in our hospitals include our hospital. Age > 70 years and GCS score < 9 at admission were predictors of mortality in patients with GA poisoning, according to Hsiao JT et al [22]. However, there was no significant difference in age, PCO2 level, or GCS score compared to published research, which may be explained by the formulation of GA herbicides differing in different countries[14]. Lee [23] also found that PCO2 level, mechanical ventilator application, and the use of vasopressors could be valuable prognostic information for evaluating the severity and mortality of patients with GA poisoning.
In our study, we observed a significant difference in initial serum ammonia, CK-MB, and the amount of GA ingested between the unfavorable outcomes group (28 patients) and the favorable outcomes group. The gold standard for the diagnosis of poisoning is poison detection, in the meaning time, the specificity of the amount of ingestion in our research was the highest. Unfortunately, most hospitals in our country do not have the capability to perform this test, and the ingestion amount of GA provided by patients. The effective of this variable in predicting prognosis was limited. Although Previous studies[1, 24] have suggested that serum ammonia levels may be a potential indicator of severe symptoms, but these findings were based on case reports with limited statistical power.
In our study, initial serum ammonia levels were higher in the unfavorable outcomes group (1.122; 1.065–1.183, P < 0.001). Additional, the levels of CK-MB were also higher in the unfavorable outcomes group (1.102; 1.020–1.198 P = 0.013). The increase of CK-MB was associated with various reason, such as myocardial injury, muscle injury, craniocerebral damage, seizure and so on. The reason that GA poisoning patients with a increased CK-MB might be related to the myocardial damage. Kim [25] published a case report about a various clinical aspects associated with cardiotoxicity after GA poisoning. In our research, 3 patients died of Cardiovascular cause(two patients suffered cardiac arrest and one patient died of cardiogenic shock). This reminded us to pay attention to cardiotoxicity as well as neurotoxicity of GA poisoning. The intracellular accumulation of ammonia is a result of the inhibition of glutamine synthesizer. Hyperammonemia destroys cell membranes and inhibits protein synthesis. Multiple logistic regression analyses indicated the initial serum ammonia level and the level of CK-MB were the indicators of unfavorable outcomes. The optimal point of initial serum ammonia was 62.7mmol/L (sensitivity:85.7%, specificity:87.2%, and area under the curve was 0.873) and the optimal level of CK-MB was 43.5U/L (sensitivity:42.9%, specificity:87.2%, and area under the curve was 0.680) according to ROC curves and the youden index.
There are several limits in our study. Firstly, this study was a single center retrospective study. Secondly, we collected the data of 67 patients, which the sample size was not large. Additional, we did not perform cranial MRI scan in patients with hyperammonemia.