Analysis of Prognostic Factors in Patients with Paraquat Poisoning for Better Therapy Regimen

Paraquat poisoning is associated with very high mortality rate and extremely dicult to manage due to lack of antidotes. The purpose of this study was to identify prognostic factors after paraquat poisoning and discuss the ecacy of current therapy regimen. Methods In this retrospective study, 211 paraquat poisoning cases admitted to the First Aliated Hospital, School of Medicine, Zhejiang University between 1 June 2010 and 30 April 2016 were enrolled. The demographic characteristic, medical records of clinical features, laboratory parameters, therapy regimen and the prognosis were analyzed. The Cox regression survival demonstrated function at or acute were Higher


Introduction
Paraquat (N, N-dimethyl-4, 4'-bipyridinium dichloride; PQ) is a widely used herbicide in developing countries for weed elimination. It exerts herbicidal activity through redox-cycling and leads to formation of reactive oxygen species (ROS). ROS interacts with the unsaturated lipids on cellular membranes, resulting in destruction of plant organelles and inevitably leading to cell death [1].
However, PQ has been demonstrated to be highly toxic to human and animal, and PQ poisoning accounts for a high mortality rate each year either by accidental or intentional exposure for suicide [2,3]. PQ enters into the body mainly by ingestion, dermal exposure or inhalation. Even a teaspoon of concentrated PQ can result in death, which is usually due to respiratory failure within days up to a month after exposure.
Besides the lung, PQ also damages the kidneys, liver, heart and central nervous system and leads to multiple organ dysfunction [4,5].
Unfortunately, there are neither effective antidotes nor widely accepted guidelines for treatment of PQ poisoning up till now. The therapy regimen varies from supportive care alone to various combinations of glucocorticoid, cyclophosphamide, anti-oxidants and immunosuppression therapy [6,7]. The activated charcoal hemoperfusions are used to increase the elimination of PQ from plasma [8][9][10][11]. Even though, the overall mortality rate are still higher than 50%, ranging from 60-80% [12,13].
Therefore, more effective therapies are urgently required. Although researchers have been dedicated to exploring new insights from numerous animal studies, conclusions from large sample clinical studies are more practical. Due to the di culty of designing perspective clinical trials in suicide population, this retrospective study enrolled 211 PQ poisoning patients in the aim of identifying prognostic factors and determining the e cacy of therapy regimen from one single center in China.

Ethics statement
This study was approved by the Ethics Committee of the First A liated Hospital, College of Medicine, Zhejiang University (Reference Number 2016 − 273). Because all information obtained from the medical records was kept con dential, the review board of the Ethics Committee stated that written consent from patients was not required. The study was conducted according to the principles expressed in the Declaration of Helsinki. Authors had access to information that could identify individual participants during or after data collection.

Study population
Between 1 June 2010 and 30 April 2016, there were 224 PQ poisoning patients admitted to the medical wards of the First A liated Hospital, School of Medicine, Zhejiang University. We excluded eight cases who ate the vegetable or fruits sprayed by PQ and ve cases who exposed to PQ by skin accidentally. This retrospective study enrolled 211 patients, they were divided into the survivor group (n = 117) and non-survivor group (n = 94) according to the prognosis ( Fig. 1 Enrollment and grouping of PQ poisoning patients). The patients were followed up till 30 April 2016.

Clinical variables
Demographic characteristics, the medical records of clinical features (the amount of PQ ingested, whether combined with alcohol drink and time interval from exposure to admission), laboratory parameters (the blood oxygen saturation, arterial oxygen pressure, serum creatinine level, glomerular ltration rate and serum uric acid level at admission, the variation trend of renal function and pulmonary CT scan), therapy regimen (gastric lavage, frequency of hemoperfusion, dose of cyclophosphamide, methylprednisolone, rapamycin, vitamin C and aspirin) and prognosis (survival and mortality) in 211 patients were retrospectively analyzed. The patients who survived at discharge were followed up by telephone.

Statistical analysis
Results were presented as the mean ± standard deviation and percentage. Chi-square test, Fisher exact test, independent t-tests and Mann-Whitney U tests were used for comparisons between groups where appropriate. The Cox regression survival model was applied to determine the independent prognostic factors. The hazard ratios (HRs) for mortality and 95% con dence intervals (CIs) were calculated. All the data were analyzed by SPSS statistical software (version 16.0; SPSS, Inc., Chicago, IL). All p-values were two-tailed and a p-value < 0.05 was considered signi cant.

Results
The mean age of 211 cases was 35.85 ± 15.29   Differences between the survivor and non-survivor group There were no signi cant differences in the distribution of most baseline variables (the gender, time interval between ingestion and admission, gastric lavage percentage, blood oxygen saturation, arterial oxygen pressure and serum uric acid level at admission) between the two groups. The majority were treated with hemoperfusion, intravenous methylprednisolone (MP), cyclophosphamide (CP) and vitamin C (VC). The pulse use of MP, pulse use of CP and intravenous use of VC in PQ poisoning patients were 97.16%, 82.46% and 93.84%, respectively. In order to block the redox cycling and pulmonary brosis, less than 50% patients were treated with oral aspirin (34.12%) and rapamycin (23.70%).
Twelve Patients who ingested PQ combined with alcohol had a higher survival rate (83.33% vs 53.77%, p = 0.045). The patients in the survivor group ingested less amount of PQ, presented with lower serum creatinine level and higher glomerular in ltration rate at admission, developed lower incidence of acute kidney injury and pulmonary CT deterioration. As to the therapy regimen, the survivors were treated with higher dosage of methylprednisolone, aspirin and rapamycin. The frequency of hemoperfusion was more in the survivor group (Table 1).  Discussion Management of PQ poisoning is a medical challenge due to its high toxicity without effective antidote. PQ is rapidly distributed to lung, liver, kidney and muscle upon ingestion and selectively accumulates in the lungs, leading to irreversible pulmonary brosis and resulting in death within days up to a month. The clinical manifestations depend upon the quantity ingested. Large amounts of liquid concentrate (> 50-100 ml of 20% ion w/v) could result in fulminant organ failure and death within several hours to a few days, while smaller quantities could be harmful to the key target organs (kidneys and lungs) and develop over the next 2-6 days with the mortality rate still over 50% [14].
In the present study, our retrospective analysis revealed that the mortality rate was 44.55%, a little better than the previous reports. Most baseline variables had no signi cant differences between the survivor and non-survivor group. However, patients in the survivor group ingested less amount of PQ, which is consistent with previous reports [14]. We also found patients presented with lower serum creatinine level and higher glomerular in ltration rate at admission, developed lower incidence of acute kidney injury in the survivor group (Table 1). Cox regression survival analysis revealed that patients with abnormal renal function at admission or developed AKI after admission were high-risk population for mortality (  [15]. In their study, AKI developed fully at the fth day after PQ ingestion and normalized within 3 weeks without exception; Serum uric acid level could be a marker for mortality and acute kidney injury in patients with acute PQ intoxication [15,16]. We could not characterize the evolution of AKI in present study because six patients who had impaired renal function at discharge did not have further renal function examinations. We could not nd the association between serum uric acid level and AKI development or prognosis, but Cox regression analysis demonstrated that the amount of PQ ingested was the only independent risk factor for the development of AKI [HR 1.042, 95% CI (1.016-1.069), p = 0.002].
The percentage of abnormal pulmonary CT imaging in the survivor group was lower than the non-survivor group (41.88% vs 57.45%, p = 0.025) in our study, the lesions were consisted of pulmonary segment involvement, effusion, consolidation and brosis, or rapid lesion progression. Previous study reported that CT imaging could be a prognostic indicator for patients with pulmonary injury from acute PQ poisoning [17]. However, Cox regression analysis in the present study could not identify it as an independent risk factor for the prognosis. Further study should put emphasis on uni ed grouping of pulmonary lesion and identify characteristic early features.
Another interesting nding was that 12 Patients who ingested PQ combined with alcohol had a higher survival rate than those ingested PQ alone. PQ induced redox cycling rapidly oxidizes NADPH and leads to secondary changes on cellular metabolism and impairs defenses against oxidative stress [18]. Previous animal studies and medical case reports drew different conclusions about the impact of ethanol on PQ poisoning. Ethanol may decrease PQ toxicity partly through competitive consumption of NADPH and oxygen during the metabolism process [19][20][21][22][23]. The acute and chronic ethanol ingestion may have different impact on PQ metabolism. Further detailed studies are needed.
As to the therapy regimen, the survivors in the present study were treated with higher dosage of methylprednisolone, aspirin and rapamycin. The frequency of hemoperfusion was more in the survivor group. In the past decades, attempts to reduce absorption by gastric lavage, administration of Fueller's earth and skin decontamination have been used routinely [24]. Early hemoperfusion is suggested in the aim of eliminating plasma PQ [9,25]. A combination of glucocorticoid and cyclophosphamide was reported to be bene cial [26]. The survival bene t of additional immunosuppressive treatment has been demonstrated in the combination of methylprednisolone, cyclophosphamide and daily dexamethasone [6]. Intravenous anti-oxidants such as N-Acetylcysteine, L-Glutathione, vitamin C, vitamin E and thioctic acid have been used with various success [7,27]. Our previous animal study revealed that rapamycin has signi cant inhibitory effects on progressive pulmonary brosis in the PQ intoxication mice model which may be partly ascribed to the inhibition of TGF-β1 [28], however, the results in clinical settings differed [29,30]. Some other potential agents, such as docosahexaenoic acid [31], naringin [32] and lysine acetylsalicylate [33] have been shown to ameliorate PQ-induced pulmonary brosis in animal models. Our results demonstrated positive effects of the therapy regimen consisted of cyclophosphamide, methylprednisolone, vitamin C, aspirin and rapamycin combined with hemoperfusion, higher dose of methylprednisolone and aspirin were protective factors for survival.
Ananieva et al. reported that salicylic acid (SA) mediates tolerance in barley plants to PQ, exogenous treatment with SA could antagonize PQ toxicity via elicitation of an antioxidative response in barley plants over ten years ago [34]. Recent studies have shown that salicylates, including sodium salicylate (NaSAL) [35,36] and lysine acetylsalicylate (LAS) [33] may form complexes with PQ, prevent its toxicity through anti-in ammatory, anti-oxidant and anti-thrombogenic properties. LAS (also named Aspirin-DL-Lysine) has been shown to be a promising intravenous antidote for the treatment of PQ poisoning in animal studies [33]. Previous study showed that the addition of high-dose vitamin C to the treatment can reduce the development of acute kidney injury and mortality in PQ poisoning patients [7]. The optimal dosage of vitamin C is unknown. Hong et al. suggested that the loading and maintenance dosages should be 2278 mg and 146 mg/h based on pharmacokinetic data in patients with PQ intoxication [27]. A much larger dose (1.5 g/kg or daily 10 g) of vitamin C recommended as the upper limit has been bene cial for the treatment of pancreatitis and advanced malignancies [37,38]. In the present study, the median daily dosage of vitamin C was 3000 mg. Further investigation is needed to determine the optimal dosage of vitamin C.
There are some limitations in present study. First, we used the amount of PQ ingestion as a variable. It's not accurate because sometimes the patient provided mistake information. We used semi-quantitative urine PQ test instead of quantitative dithionite concentration determination. As we know, 90% of the absorbed PQ is rapidly excreted unchanged in urine within 12-24 h after ingestion [39], the urine test has a good correlation between PQ concentration and intensity of the blue color formed [40]. However, due to the different time interval between ingestion and admission, we could not decide the therapy intensity according to the urine semi-quantitative results alone. Second, the dosage of aspirin was determined according to our own experience; further studies are needed to verify the conclusion and illuminate the optimal dosage.

Conclusions
In conclusion, patients who ingested larger amount of PQ and presented with abnormal renal function at admission or developed AKI after admission are high-risk population for mortality after PQ poisoning. Present study proved that either higher dose of methylprednisolone or aspirin help to improve survival. As to our knowledge, it's the rst report about the e cacy of aspirin on PQ poisoning. However, further investigations are needed to determine the optimal dosage. Kaplan-Meier survival analysis between two groups according to the AKI incidence after admission. Logrank test, p<0.001 Kaplan-Meier survival analysis between two groups according to the GFR at admission. Log-rank test, p<0.001

Figure 3
Survival according to different renal function grouping