- Email: [email protected]
Rapid Analysis of Plasma Paraquat Using Sodium Dithionite As a Predictor of Outcome in Acute Paraquat Poisoning
CLINICAL INVESTIGATION
Rapid Analysis of Plasma Paraquat Using Sodium Dithionite As a Predictor of Outcome in Acute Paraquat Poisoning Ja-Ryong Koo, MD, Jong-Woo Yoon, MD, Sang-Jin Han, MD, Myung-Jin Choi, MD, In-Il Park, MD, Young-Ki Lee, MD, Sung-Gyun Kim, MD, Ji-Eun Oh, MD, Jang-Won Seo, MD, Hyung-Jik Kim, MD and Jung-Woo Noh, MD
Abstract: Background: Paraquat poisoning can be lethal, and aggressive treatments might have little or no effect on severely poisoned patients. Accordingly, a convenient prognostic test is necessary to guide therapy for acute paraquat poisoning. Sodium dithionite reduces paraquat to a blue radical form in alkaline plasma with a paraquat detection sensitivity of 2.0 mg/L, which is a 100% lethal concentration at 10 to 12 hours postingestion. The prognostic utility of this simple reaction was examined prospectively. Methods: Of 233 paraquatpoisoned patients, who were taken to the hospital within 12 hours after ingestion, the plasma samples obtained on arrival were tested using the sodium dithionite reaction. Standard saline containing 2.0 mg/L paraquat was used as the positive control. The test result was interpreted as being positive when the plasma test yielded a blue color darker than that of the positive control. The effects of aggressive treatment, including cyclophosphamide pulse and continuous venovenous hemofiltration, were evaluated retrospectively. Results: The discharge survival rate was 41.6% (97 of 233). Ninety-seven of 142 patients with negative or equivocal plasma dithionite test survived. However, all 91 patients with positive plasma dithionite test died of multiorgan failure. Cyclophosphamide and/or continuous venovenous hemofiltration could not improve survival. Conclusions: In this single-center study, a positive plasma dithionite test was associated with 100% mortality, despite aggressive treatment. In contrast, negative or equivocal tests were associated with a 68% survival rate. It is believed that after further verification, this test can be used to guide therapy and predict the outcomes of patients suffering acute paraquat poisoning. Key Indexing Terms: Paraquat poisoning; Sodium dithionite; Plasma paraquat level; Prognostic marker. [Am J Med Sci 2009; 338(5):373–377.]
P
araquat (1,1⬘-dimethyl-4,4⬘-bipyridium dichloride) has been marketed since the mid-1960s as a contact herbicide and desiccant with low chronic toxicity because of its rapid deactivation on soil contact.1 Although proven to be effective and safe in routine agricultural use, paraquat has received notoriety throughout the world because of the large number of human deaths that have occurred after its ingestion.1 Gramoxone (Syngenta, Basel, Switzerland) is the most common trade name for paraquat, but the herbicide is also sold under many different trade names (Crisquat, Dextrone X, Esgram, Preeglone, Priglone, and Weedol) by several different companies. Because of its easy availability and high fatality, paraquat From the Department of Internal Medicine, Hallym Kidney Research Institute, College of Medicine, Hallym University, Chuncheon, Korea. Submitted January 5, 2009; accepted in revised form June 25, 2009. Drs. Koo and Yoon have equally contributed to this work. Correspondence: Ja-Ryong Koo, MD, Division of Nephrology, Department of Internal Medicine, Chuncheon Sacred Heart Hospital, Chuncheon, Gangwon-Do, Korea 200-704 (E-mail: [email protected]).
poisoning with suicidal intent is a significant clinical problem in parts of the Asia, Pacific region, and the Caribbean.1,2 In Korea, paraquat accounts for most fatal poisonings, with 500 or more associated deaths per year.3 Paraquat poisoning is characterized by severe free radical injury, which causes early multiorgan failure, including kidney and late pulmonary fibrosis with respiratory failure. The management of paraquat poisoning is directed at removing paraquat from the body and preventing the radical-induced organ damage using antiinflammatory agents.1,2 Although aggressive treatment using continuous venovenous hemofiltration (CVVH) and/or immunosuppressive agents is successful in some patients with mild poisoning,2,3 there is no consensus on its efficacy, and aggressive treatment might be futile in patients with severe paraquat poisoning who have no hope of survival.3 The severity and outcome of paraquat poisoning is determined primarily by the dose ingested. However, the quantity ingested is often difficult to determine, particularly in intoxicated patients or patients presenting with confusion. In addition, vomiting after ingestion and the gastrointestinal adsorption of paraquat by food can reduce the actual amount of paraquat absorbed.1 Measuring the plasma paraquat concentration is of unquestionable value in both assessing the prognosis and deciding whether to use or withhold aggressive therapy.4 A variety of spectrophotometric and chromatographic techniques with a sensitivity limit ⬍0.1 mg/L have been used to measure the plasma paraquat concentration.4 However, most methods described are limited by several factors, such as the availability of equipment and the clinical demand in emergency situations.3,4 Proudfoot et al5 reported that patients whose plasma paraquat concentrations do not exceed 2.0 mg/L at 10 hours postingestion are likely to survive. In the survival curve produced by Lheureux and Ekwall,6 a blood paraquat level of 2.0 mg/L was a 100% lethal concentration at 10 to 12 hours postingestion. Moreover, Gil et al,7 who evaluated the association between the plasma paraquat levels and outcomes in 375 paraquat-poisoned patients, reported that no patients with plasma paraquat levels of ⬎2.0 mg/L 4 hours postingestion could survive. Therefore, all patients with a plasma paraquat level ⬍2.0 mg/L have the potential for recovery with vigorous treatment irrespective of the accurate plasma level if they are admitted to the hospital within 10 to 12 hours after ingestion. However, for the same reason, aggressive therapy has proven to be futile if the plasma paraquat levels are ⬎2.0 mg/L, particularly after 4 hours postingestion. Accordingly, a simple and convenient plasma paraquat detection test with a sensitivity limit of approximately 2.0 mg/L is urgently required as a prognostic marker. In alkaline media, paraquat forms a blue radical ion1 in the presence of dithionite, which can be used to detect paraquat in various media. This semi-quantitative reaction is simple and rapid
The American Journal of the Medical Sciences • Volume 338, Number 5, November 2009
373
Koo et al
and has been used widely to detect paraquat in urine with a detection limit of 1 to 2 mg/L.4 In our preliminary study using standards containing 1.0 to 3.0 mg of paraquat per liter of saline, a clear bluish color change began at 2.0 mg/L, which corresponded to the target sensitivity limit for paraquat detection. In view of these considerations, sodium dithionite was applied to plasma to detect plasma paraquat, and the prognostic significance and clinical utility of this test in determining whether to use aggressive or conservative therapy for paraquat poisoning in emergency situations was examined prospectively. The effect of aggressive treatment on the patients’ survival, including cyclophosphamide pulse and CVVH, was also evaluated retrospectively based on the plasma sodium dithionite test results.
METHODS This study was performed between May 1, 1999, and March 30, 2007. Patients were enrolled from a tertiary care regional university hospital (Hallym University Hospital, Chuncheon, Korea). The patients fulfilling the following inclusion criteria were enrolled in this study: 18 years of age or older, had ingested 20% liquid paraquat concentrate with suicidal intent, and presented to an emergency unit within 12 hours after paraquat ingestion. The exclusion criteria included patients with accidental paraquat ingestion, patients with diquat poisoning, and patients presenting to the hospital ⬎12 hours after paraquat ingestion. In all patients, urine and plasma samples taken on arrival at the emergency room were tested using the sodium dithionite reaction. In our laboratory, an alkaline sodium dithionite solution was prepared immediately by adding 100 mg of sodium dithionite (Hayashi Pure Chemical, Japan) to 10 mL of 2 M sodium hydroxide (the reagent is stable for approximately 2 hours). An aliquot sample (200 L) of this solution was added to 2 mL of urine and plasma with the negative and positive control. Standards containing 0 and 2.0 mg paraquat per liter of saline were used as the negative and positive controls. The test results were interpreted as being positive when the urine or plasma tests yielded a blue color that was darker than the positive control by the naked eye. Plasma with a faint blue color was classified as equivocal if the intensity of the bluish color change was similar to that of the positive control. Two experienced observers interpreted the results. The results were reported as equivocal in cases of disagreement on the plasma tests between the 2 observers. The entire procedure was completed within 10 minutes for a plasma sample and within 5 minutes for a urine sample. To measure the reproducibility of the plasma paraquat test, 80 tests were repeated using the same plasma samples. The serum aspartate transaminase, alanine transaminase, and creatinine levels were measured using standard techniques. To prevent the absorption of paraquat from the gastrointestinal tract, activated charcoal or Fuller’s earth added to 20% mannitol was administered through a nasogastric tube after gastric lavage with tap water. All patients then received 1 to 3 courses of 6-hour charcoal hemoperfusion therapy (Adsorba; Gambro, Sweden; charcoal carbon, 300 g) using an internal jugular or femoral dual lumen catheter. The number of charcoal hemoperfusion treatments was determined by the results of the urine sodium dithionite test. In cases of negative urine tests, only 1 hemoperfusion was attempted, whereas 2 to 3 hemoperfusions were used in cases with a positive urine test. In 90 patients with acute renal failure, CVVH was started after hemoperfusion using a Prisma or AK-10 blood pump (Gambro). The blood flow rate for CVVH was 100 to 150 mL/min. The effluent filtration rate was 1500 to 2000 mL/hour, and the
374
duration of CVVH, replacement fluid, and net ultrafiltration rates varied according to the clinical status of the patients. To reduce oxygen radical injury, intravenous steroid (methylprednisolone 1 g/d or dexamethasone 10 mg every 6 hours for 3 days) was administered to all patients. The steroid was tapered slowly over several weeks. For all patients with multiorgan failure, standard treatments, including volume trial, vasopressor, and ventilatory support, were attempted. Based on the report by Lin et al,8 which demonstrated improved survival with a cyclophosphamide and corticosteroid pulse, intravenous cyclophosphamide (0.75–1 g/d for 3 days) was added to prevent delayed lung fibrosis in 137 patients. The outcome measurement was survival at discharge from hospital. To examine the prognostic significance of the sodium dithionite test, positive and negative predictive values for the prediction of death were calculated for the plasma and urine sodium dithionite tests. The additional therapeutic benefits of the cyclophosphamide pulse and/or CVVH for patient survival were evaluated retrospectively based on the plasma sodium dithionite test results. The study protocol was approved by the Hallym University Hospital institutional review board, and all patients or their family members provided informed consent for the treatment options. Data analysis was performed using SPSS for Windows version 11.5. The differences between the groups were compared using a student t and 2 tests. All the results are reported as mean ⫾ standard deviation. A 2-tailed P value ⬍0.05 was considered significant.
RESULTS During the study period, a total of 285 patients with a diagnosis of acute paraquat poisoning were admitted to the hospital. Fifty-two patients were excluded from the evaluation for the following reasons: accidental ingestion in 7 patients; follow-up loss or transfer to another hospital in 10 patients; treatment refusal in 12 patients; and late presentation (⬎12 hours after ingestion) to the hospital in 23 patients. Accordingly, a total of 233 paraquat-poisoned patients were enrolled in this study. The total survival rate at discharge was 41.6% (97 of 233). Among the 233 patients, 40 patients were negative on the urine sodium dithionite test. All patients with negative urine sodium dithionite tests were also negative on the plasma sodium dithionite tests, and all patients survived, resulting in a 100% negative predictive value for the prediction of death. Among the 193 remaining patients with positive urine sodium dithionite tests, the plasma sodium dithionite tests were negative in 90 patients, equivocal in 12 patients, and positive in 91 patients. In the 3 subsets of patients, 51 of 90, 6 of 12, and none of 91 patients survived. Accordingly, the plasma sodium dithionite test had a 100% positive predictive value for the prediction of death. In 12 patients with an equivocal plasma sodium dithionite test, 6 patients in whom treatment could be started within 4 hours after ingestion survived. However, none of the remaining 6 patients with equivocal results, who presented to the hospital ⬎4 hours after ingestion, survived. Table 1 lists the principal characteristics of the patients and comparisons according to urine and plasma sodium dithionite tests. Because no patient with a positive plasma sodium dithionite test survived, the additional therapeutic benefits of the cyclophosphamide pulse and/or CVVH were evaluated retrospectively only in the subgroup of patients with negative or equivocal plasma sodium dithionite tests (Table 2). The analysis revealed no significant difference in mortality between the treatment groups. Volume 338, Number 5, November 2009
Plasma Dithionite Test in Paraquat Poisoning
TABLE 1. Comparisons according to urine (U)/plasma (P) dithionite test results Urine (U)/plasma (P) dithionite test result Variables
(U)Negative/(P)Negative (n ⴝ 40)
(U)Positive/(P)Negative (n ⴝ 90)
(U)Positive/(P)Equivocal (n ⴝ 12)
(U)Positive/(P)Positive (n ⴝ 91)
P
40.6 ⫾ 12.3 24 (60.0) 0.6 ⫾ 0.6 5.2 ⫾ 2.1 15 (37.5) 25 (62.5) 72 ⫾ 19
46.0 ⫾ 15.8 57 (63.3) 1.9 ⫾ 1.0 5.9 ⫾ 2.2 35 (38.9) 51 (56.7) 112 ⫾ 75
48.3 ⫾ 14.7 6 (50.0) 2.1 ⫾ 0.8 5.3 ⫾ 3.3 3 (25.0) 6 (50.0) 141 ⫾ 76
46.3 ⫾ 16.5 51 (56.0) 2.6 ⫾ 0.6 5.7 ⫾ 2.9 37 (40.7) 55 (60.4) 133 ⫾ 107
0.203 0.695 ⬍0.001 0.480 0.772 0.830 0.002
80 ⫾ 16
227 ⫾ 221
293 ⫾ 223
306 ⫾ 165
⬍0.001
2 (5.0) 33.8 ⫾ 28.6 27.8 ⫾ 19.3 10.4 ⫾ 5.5 1 (2.5) 0 (0)
40 (44.4) 48.0 ⫾ 52.1 31.2 ⫾ 27.1 13.0 ⫾ 6.8 48 (53.3) 39 (43.3)
7 (58.3) 69.3 ⫾ 54.5 54.7 ⫾ 32.2 17.3 ⫾ 6.8 6 (50.0) 6 (50.0)
74 (81.3) 87.8 ⫾ 133.1 51.5 ⫾ 60.2 15.4 ⫾ 15.2 86 (94.5) 91 (100)
⬍0.001 0.007 0.003 0.062 ⬍0.001 ⬍0.001
Age (yr) Male, n (%) Ingestion amount, mouthful Time to HP (hr) CVVH treatment, n (%) CYC pulse, n (%) Initial serum creatinine (mol/L) Peak serum creatinine (mol/L) Acute renal failurea, n (%) AST (U/L) ALT (U/L) Total bilirubin (mol/L) Hypoxiab, n (%) Mortality, n (%)
Values are expressed as mean ⫾ SD unless otherwise noted. a Acute renal failure: serum creatinine ⬎133 mol/L. b Hypoxia: PaO2 ⬍60 mmHg. HP, hemoperfusion; CVVH, continuous venovenous hemofiltration; CYC, cyclophosphamide; AST, aspartate transaminase; ALT, alanine transaminase.
The reproducibility of the plasma sodium dithionite test was 98% for positive cases and 92% for negative or equivocal cases with a kappa score of 0.915.
DISCUSSION Although the plasma paraquat concentration seems to be the most useful marker of the severity of paraquat poisoning,
measurements of the plasma paraquat level are not readily available in most hospitals. Moreover, most currently available methods require experience and rigorous quality control to provide reliable results.2,4 Most other prognostic indicators also require blood gas analysis and biochemical tests, such as potassium, bicarbonate, and creatinine concentrations, which may require several hours to obtain results and have relatively
TABLE 2. Comparisons according to treatment modalities in subgroup of patients with negative or equivocal plasma sodium dithionite test Treatment modalities Variables Age (yr) Male, n (%) Ingestion amount, mouthful Time to HP (hr) Initial serum creatinine (mol/L) Peak serum creatinine (mol/L) Acute renal failurea, n (%) AST (U/L) ALT (U/L) Total bilirubin (mol/L) Hypoxiab (%) Mortality (%)
HP only (n ⴝ 35)
HP ⴙ CVVH (n ⴝ 25)
HP ⴙ CYC (n ⴝ 54)
HP ⴙ CVVH ⴙ CYC (n ⴝ 28)
P
48.4 ⫾ 13.2 24 (68.6) 1.7 ⫾ 0.9 5.3 ⫾ 2.1 108 ⫾ 88 178 ⫾ 195 10 (28.6) 41.4 ⫾ 31.1 30.2 ⫾ 19.2 12.8 ⫾ 5.6 16 (45.7) 11 (31.4)
43.7 ⫾ 13.4 15 (60.0) 1.4 ⫾ 1.1 5.3 ⫾ 1.6 89 ⫾ 48 169 ⫾ 149 9 (36.0) 57.2 ⫾ 50.4 37.6 ⫾ 31.5 10.3 ⫾ 4.1 11 (44.0) 8 (32.0)
44.2 ⫾ 14.4 33 (61.1) 1.5 ⫾ 1.1 6.0 ⫾ 2.7 96 ⫾ 47 178 ⫾ 170 19 (35.2) 40.9 ⫾ 43.9 29.9 ⫾ 27.0 13.2 ⫾ 8.2 15 (27.8) 17 (31.5)
41.6 ⫾ 18.7 15 (53.6) 1.3 ⫾ 0.8 5.7 ⫾ 2.2 119 ⫾ 86 256 ⫾ 280 11 (39.3) 49.3 ⫾ 65.2 32.9 ⫾ 27.9 13.3 ⫾ 6.0 13 (46.4) 9 (32.1)
0.315 0.682 0.296 0.419 0.456 0.292 0.834 0.490 0.657 0.277 0.218 0.969
Values are expressed as mean ⫾ SD unless otherwise noted. a Acute renal failure: serum creatinine ⬎133 mol/L. b Hypoxia: PaO2 ⬍60 mmHg. HP, hemoperfusion; CVVH, continuous venovenous hemofiltration; CYC, cyclophosphamide; AST, aspartate transaminase; ALT, alanine transaminase.
© 2009 Lippincott Williams & Wilkins
375
Koo et al
poor discriminative powers for the prediction of survival or death.2,9 –12 Therefore, a simple and accurate prognostic marker is needed to help clinicians predict the prognosis and guide initial therapy. This study showed that a plasma/urine sodium dithionite test at admission can provide clinically important information to determine the severity and predict the prognosis of paraquat poisoning. In this study, the target sensitivity for plasma paraquat detection was as high as 2.0 mg/L, which obviates complex and time-consuming paraquat extraction or plasma protein precipitation steps.4 The entire procedure was completed within 10 minutes and could be performed easily without any equipment in emergency situations. The management of paraquat poisoning is directed at removing paraquat from the gastrointestinal tract, increasing its excretion from the blood, reducing radical-induced tissue injury with antioxidants, and preventing pulmonary damage with antiinflammatory drugs.1 Lin et al8 reported that an intravenous cyclophosphamide pulse might be effective in treating patients with pulmonary fibrosis due to paraquat poisoning. In a previous study,3 it was shown that the addition of prophylactic CVVH can allow for the successful management of paraquat-poisoned patients within the time where acute multiorgan failure can lead to immediate death caused by circulatory collapse. Therefore, the combination of CVVH, which prevents early death from multiorgan failure, and pulse cyclophosphamide, which prevents late pulmonary fibrosis, might be effective therapeutic options in paraquat poisoning. However, the efficacy of those intensive treatments, which are not without risk, has not been verified using a validated prognostic system. It is important to predict the prognosis or response to aggressive treatment and advise the patients and relatives of the likely outcomes. According to the results of this study, intensive treatment, including cyclophosphamide pulse and/or CVVH, had no therapeutic effect on patients with positive plasma sodium dithionite tests. Therefore, it is reasonable to use an intensive form of treatment preferentially in the subgroup of patients with negative plasma sodium dithionite tests. However, in this study, there was also no difference in mortality according to treatment modality in the subgroup of patients with negative plasma sodium dithionite tests. This might be due to the retrospective nature of the analysis. Therefore, a prospective study will be needed to evaluate the therapeutic effects of a cyclophosphamide pulse and/or CVVH in subgroups of patients with negative plasma sodium dithionite tests. There are some limitations in this study. The test is a color read, semi-qualitative test that uses the naked eye. Ac-
cordingly, rigorous standardization of the test will be needed before its clinical applications can be accepted. In our laboratory, 2 observers interpreted the results to reduce the likelihood of subjectivity in the interpretation of plasma dithionite tests, and any disagreement between the 2 observers was reported as equivocal. Moreover, the reproducibility of the plasma sodium dithionite test was 98% for the positive cases and 92% for the negative or equivocal cases with a kappa score of 0.915, which supports reliability of this test as a prognostic marker, particularly in definitely positive cases. Another limitation is that there might be cross-reactions with compounds such as diquat and interference from a high bilirubin level. However, in this study, diquat-poisoned patients and paraquat-poisoned patients, who presented to the hospital ⬎12 hours after paraquat ingestion, were excluded. Because liver failure usually begins 1 to 3 days after paraquat ingestion, hyperbilirubinemia induced by liver failure will rarely develop within 12 hours after paraquat ingestion. In this study, only 2 patients had baseline hyperbilirubinemia (plasma bilirubin ⬎2.0 mg/dL) at admission. According to the survival curve,5,6 which relates the outcome to the plasma paraquat concentration at admission and ingestion to the sampling interval, a small portion of patients with plasma paraquat concentrations ⬎2.0 mg/L have a chance of survival if treatment is begun within 2 to 4 hours after ingestion. Indeed, Gil et al7 reported that the upper limit of the plasma paraquat level in survivors was 2.64 mg/L at 3 hours. In conformity with the above-mentioned studies, 6 survivors of 12 patients with an equivocal plasma sodium dithionite test presented to the hospital within 4 hours of ingestion. This suggests that intensive treatment can be attempted with some success in select patients with a positive plasma sodium dithionite test if they present to the hospital within 2 to 4 hours after paraquat ingestion.
CONCLUSION The plasma sodium dithionite test is a simple and reliable prognostic marker with a 100% positive predictive value for the prediction of death in paraquat-poisoned patients who present to the hospital ⬎4 hours after ingestion. Aggressive treatment, including a cyclophosphamide pulse and/or CVVH, might have no therapeutic effect on patients with definitely positive plasma sodium dithionite tests. A relatively high survival rate can be expected in patients with negative plasma sodium dithionite tests. A treatment algorithm is suggested based on the plasma and urine sodium dithionite tests (Figure 1). However, considering the semi-quantitative nature of this test performed at a
Urine sodium dithionite test
Negative
Positive (or Anuric)
Minimal poisoning : 100% survival
Plasma sodium dithionite test
: Supportive care (± Hemoperfusion) Negative or Equivocal
Positive
Mild to moderate poisoning
Severe poisoning
: 50-60% survival
: 0% survival especially in patients
: May require intensive treatment (Hemoperfusion/Hemofiltration)
FIGURE 1. Suggested therapeutic approach for paraquat-poisoned patients based on the sodium dithionite test results.
presenting to hospital > 4hrs post-ingestion : Supportive care only
(Steroid/Cyclophosphamide)
376
Volume 338, Number 5, November 2009
Plasma Dithionite Test in Paraquat Poisoning
single institution, further confirmatory studies with rigorous standardization will be needed before this test can be used as a therapeutic guideline in the decision to use or withhold aggressive therapy. Until this test has been quantified and verified, it can be used to counsel victims and/or their families regarding the prognosis in an individual case. ACKNOWLEDGMENT The authors thank the staff of the clinical laboratory of the Chuncheon Sacred Heart Hospital for their assistance. REFERENCES 1. Dinis-Oliveira RJ, Duarte JA, Sa´nchez-Navarro A, et al. Paraquat poisonings: mechanisms of lung toxicity, clinical features, and treatment. Crit Rev Toxicol 2008;38:13–71. 2. Eddleston M, Wilks MF, Buckley NA. Prospects for treatment of paraquat-induced lung fibrosis with immunosuppressive drugs and the need for better prediction of outcome: a systematic review. QJM 2003;96:809 –24. 3. Koo JR, Kim JC, Yoon JW, et al. Failure of continuous venovenous hemofiltration to prevent death in paraquat poisoning. Am J Kidney Dis 2002;39:55–9. 4. Braithwaite RA. Emergency analysis of paraquat in biological fluids. Hum Toxicol 1987;6:83– 6.
© 2009 Lippincott Williams & Wilkins
5. Proudfoot AT, Stewart MS, Levitt T, et al. Paraquat poisoning: significance of plasma-paraquat concentrations. Lancet 1979;2:330 –2. 6. Lheureux P, Ekwall B. Time-related lethal blood concentrations from acute human poisoning of chemicals (The MEMO database) Part 2: the monographs. No. 25, Paraquat 1997. Available at: http://www. cctoxconsulting.a.se/25_paraquat.pdf. Accessed October 1, 2008. 7. Gil HW, Kang MS, Yang JO, et al. Association between plasma paraquat level and outcome of paraquat poisoning in 375 paraquat poisoning patients. Clin Toxicol 2008;46:515– 8. 8. Lin JL, Leu ML, Liu YC, et al. A prospective clinical trial of pulse therapy with glucocorticoid and cyclophosphamide in moderate to severe paraquat-poisoned patients. Am J Respir Crit Care Med 1999; 159:357– 60. 9. Suzuki K, Takasu N, Arita S, et al. A new method for predicting the outcome and survival period in paraquat poisoning. Hum Toxicol 1989;8:33– 8. 10. Yamaguchi H, Sato S, Watanabe S, et al. Pre-embarkment prognostication for acute paraquat poisoning. Hum Exp Toxicol 1990;9:381– 4. 11. Ragoucy-Sengler C, Pileire B. A biological index to predict patient outcome in paraquat poisoning. Hum Exp Toxicol 1996;15:265– 8. 12. Hong S-Y, Yang D-H, Hwang K-Y. Association between laboratory parameters and outcome of paraquat poisoning. Toxicol Lett 2000;118: 53–9.
377
Rapid Analysis of Plasma Paraquat Using Sodium Dithionite As a Predictor of Outcome in Acute Paraquat Poisoning Ja-Ryong Koo, MD, Jong-Woo Yoon, MD, Sang-Jin Han, MD, Myung-Jin Choi, MD, In-Il Park, MD, Young-Ki Lee, MD, Sung-Gyun Kim, MD, Ji-Eun Oh, MD, Jang-Won Seo, MD, Hyung-Jik Kim, MD and Jung-Woo Noh, MD
Abstract: Background: Paraquat poisoning can be lethal, and aggressive treatments might have little or no effect on severely poisoned patients. Accordingly, a convenient prognostic test is necessary to guide therapy for acute paraquat poisoning. Sodium dithionite reduces paraquat to a blue radical form in alkaline plasma with a paraquat detection sensitivity of 2.0 mg/L, which is a 100% lethal concentration at 10 to 12 hours postingestion. The prognostic utility of this simple reaction was examined prospectively. Methods: Of 233 paraquatpoisoned patients, who were taken to the hospital within 12 hours after ingestion, the plasma samples obtained on arrival were tested using the sodium dithionite reaction. Standard saline containing 2.0 mg/L paraquat was used as the positive control. The test result was interpreted as being positive when the plasma test yielded a blue color darker than that of the positive control. The effects of aggressive treatment, including cyclophosphamide pulse and continuous venovenous hemofiltration, were evaluated retrospectively. Results: The discharge survival rate was 41.6% (97 of 233). Ninety-seven of 142 patients with negative or equivocal plasma dithionite test survived. However, all 91 patients with positive plasma dithionite test died of multiorgan failure. Cyclophosphamide and/or continuous venovenous hemofiltration could not improve survival. Conclusions: In this single-center study, a positive plasma dithionite test was associated with 100% mortality, despite aggressive treatment. In contrast, negative or equivocal tests were associated with a 68% survival rate. It is believed that after further verification, this test can be used to guide therapy and predict the outcomes of patients suffering acute paraquat poisoning. Key Indexing Terms: Paraquat poisoning; Sodium dithionite; Plasma paraquat level; Prognostic marker. [Am J Med Sci 2009; 338(5):373–377.]
P
araquat (1,1⬘-dimethyl-4,4⬘-bipyridium dichloride) has been marketed since the mid-1960s as a contact herbicide and desiccant with low chronic toxicity because of its rapid deactivation on soil contact.1 Although proven to be effective and safe in routine agricultural use, paraquat has received notoriety throughout the world because of the large number of human deaths that have occurred after its ingestion.1 Gramoxone (Syngenta, Basel, Switzerland) is the most common trade name for paraquat, but the herbicide is also sold under many different trade names (Crisquat, Dextrone X, Esgram, Preeglone, Priglone, and Weedol) by several different companies. Because of its easy availability and high fatality, paraquat From the Department of Internal Medicine, Hallym Kidney Research Institute, College of Medicine, Hallym University, Chuncheon, Korea. Submitted January 5, 2009; accepted in revised form June 25, 2009. Drs. Koo and Yoon have equally contributed to this work. Correspondence: Ja-Ryong Koo, MD, Division of Nephrology, Department of Internal Medicine, Chuncheon Sacred Heart Hospital, Chuncheon, Gangwon-Do, Korea 200-704 (E-mail: [email protected]).
poisoning with suicidal intent is a significant clinical problem in parts of the Asia, Pacific region, and the Caribbean.1,2 In Korea, paraquat accounts for most fatal poisonings, with 500 or more associated deaths per year.3 Paraquat poisoning is characterized by severe free radical injury, which causes early multiorgan failure, including kidney and late pulmonary fibrosis with respiratory failure. The management of paraquat poisoning is directed at removing paraquat from the body and preventing the radical-induced organ damage using antiinflammatory agents.1,2 Although aggressive treatment using continuous venovenous hemofiltration (CVVH) and/or immunosuppressive agents is successful in some patients with mild poisoning,2,3 there is no consensus on its efficacy, and aggressive treatment might be futile in patients with severe paraquat poisoning who have no hope of survival.3 The severity and outcome of paraquat poisoning is determined primarily by the dose ingested. However, the quantity ingested is often difficult to determine, particularly in intoxicated patients or patients presenting with confusion. In addition, vomiting after ingestion and the gastrointestinal adsorption of paraquat by food can reduce the actual amount of paraquat absorbed.1 Measuring the plasma paraquat concentration is of unquestionable value in both assessing the prognosis and deciding whether to use or withhold aggressive therapy.4 A variety of spectrophotometric and chromatographic techniques with a sensitivity limit ⬍0.1 mg/L have been used to measure the plasma paraquat concentration.4 However, most methods described are limited by several factors, such as the availability of equipment and the clinical demand in emergency situations.3,4 Proudfoot et al5 reported that patients whose plasma paraquat concentrations do not exceed 2.0 mg/L at 10 hours postingestion are likely to survive. In the survival curve produced by Lheureux and Ekwall,6 a blood paraquat level of 2.0 mg/L was a 100% lethal concentration at 10 to 12 hours postingestion. Moreover, Gil et al,7 who evaluated the association between the plasma paraquat levels and outcomes in 375 paraquat-poisoned patients, reported that no patients with plasma paraquat levels of ⬎2.0 mg/L 4 hours postingestion could survive. Therefore, all patients with a plasma paraquat level ⬍2.0 mg/L have the potential for recovery with vigorous treatment irrespective of the accurate plasma level if they are admitted to the hospital within 10 to 12 hours after ingestion. However, for the same reason, aggressive therapy has proven to be futile if the plasma paraquat levels are ⬎2.0 mg/L, particularly after 4 hours postingestion. Accordingly, a simple and convenient plasma paraquat detection test with a sensitivity limit of approximately 2.0 mg/L is urgently required as a prognostic marker. In alkaline media, paraquat forms a blue radical ion1 in the presence of dithionite, which can be used to detect paraquat in various media. This semi-quantitative reaction is simple and rapid
The American Journal of the Medical Sciences • Volume 338, Number 5, November 2009
373
Koo et al
and has been used widely to detect paraquat in urine with a detection limit of 1 to 2 mg/L.4 In our preliminary study using standards containing 1.0 to 3.0 mg of paraquat per liter of saline, a clear bluish color change began at 2.0 mg/L, which corresponded to the target sensitivity limit for paraquat detection. In view of these considerations, sodium dithionite was applied to plasma to detect plasma paraquat, and the prognostic significance and clinical utility of this test in determining whether to use aggressive or conservative therapy for paraquat poisoning in emergency situations was examined prospectively. The effect of aggressive treatment on the patients’ survival, including cyclophosphamide pulse and CVVH, was also evaluated retrospectively based on the plasma sodium dithionite test results.
METHODS This study was performed between May 1, 1999, and March 30, 2007. Patients were enrolled from a tertiary care regional university hospital (Hallym University Hospital, Chuncheon, Korea). The patients fulfilling the following inclusion criteria were enrolled in this study: 18 years of age or older, had ingested 20% liquid paraquat concentrate with suicidal intent, and presented to an emergency unit within 12 hours after paraquat ingestion. The exclusion criteria included patients with accidental paraquat ingestion, patients with diquat poisoning, and patients presenting to the hospital ⬎12 hours after paraquat ingestion. In all patients, urine and plasma samples taken on arrival at the emergency room were tested using the sodium dithionite reaction. In our laboratory, an alkaline sodium dithionite solution was prepared immediately by adding 100 mg of sodium dithionite (Hayashi Pure Chemical, Japan) to 10 mL of 2 M sodium hydroxide (the reagent is stable for approximately 2 hours). An aliquot sample (200 L) of this solution was added to 2 mL of urine and plasma with the negative and positive control. Standards containing 0 and 2.0 mg paraquat per liter of saline were used as the negative and positive controls. The test results were interpreted as being positive when the urine or plasma tests yielded a blue color that was darker than the positive control by the naked eye. Plasma with a faint blue color was classified as equivocal if the intensity of the bluish color change was similar to that of the positive control. Two experienced observers interpreted the results. The results were reported as equivocal in cases of disagreement on the plasma tests between the 2 observers. The entire procedure was completed within 10 minutes for a plasma sample and within 5 minutes for a urine sample. To measure the reproducibility of the plasma paraquat test, 80 tests were repeated using the same plasma samples. The serum aspartate transaminase, alanine transaminase, and creatinine levels were measured using standard techniques. To prevent the absorption of paraquat from the gastrointestinal tract, activated charcoal or Fuller’s earth added to 20% mannitol was administered through a nasogastric tube after gastric lavage with tap water. All patients then received 1 to 3 courses of 6-hour charcoal hemoperfusion therapy (Adsorba; Gambro, Sweden; charcoal carbon, 300 g) using an internal jugular or femoral dual lumen catheter. The number of charcoal hemoperfusion treatments was determined by the results of the urine sodium dithionite test. In cases of negative urine tests, only 1 hemoperfusion was attempted, whereas 2 to 3 hemoperfusions were used in cases with a positive urine test. In 90 patients with acute renal failure, CVVH was started after hemoperfusion using a Prisma or AK-10 blood pump (Gambro). The blood flow rate for CVVH was 100 to 150 mL/min. The effluent filtration rate was 1500 to 2000 mL/hour, and the
374
duration of CVVH, replacement fluid, and net ultrafiltration rates varied according to the clinical status of the patients. To reduce oxygen radical injury, intravenous steroid (methylprednisolone 1 g/d or dexamethasone 10 mg every 6 hours for 3 days) was administered to all patients. The steroid was tapered slowly over several weeks. For all patients with multiorgan failure, standard treatments, including volume trial, vasopressor, and ventilatory support, were attempted. Based on the report by Lin et al,8 which demonstrated improved survival with a cyclophosphamide and corticosteroid pulse, intravenous cyclophosphamide (0.75–1 g/d for 3 days) was added to prevent delayed lung fibrosis in 137 patients. The outcome measurement was survival at discharge from hospital. To examine the prognostic significance of the sodium dithionite test, positive and negative predictive values for the prediction of death were calculated for the plasma and urine sodium dithionite tests. The additional therapeutic benefits of the cyclophosphamide pulse and/or CVVH for patient survival were evaluated retrospectively based on the plasma sodium dithionite test results. The study protocol was approved by the Hallym University Hospital institutional review board, and all patients or their family members provided informed consent for the treatment options. Data analysis was performed using SPSS for Windows version 11.5. The differences between the groups were compared using a student t and 2 tests. All the results are reported as mean ⫾ standard deviation. A 2-tailed P value ⬍0.05 was considered significant.
RESULTS During the study period, a total of 285 patients with a diagnosis of acute paraquat poisoning were admitted to the hospital. Fifty-two patients were excluded from the evaluation for the following reasons: accidental ingestion in 7 patients; follow-up loss or transfer to another hospital in 10 patients; treatment refusal in 12 patients; and late presentation (⬎12 hours after ingestion) to the hospital in 23 patients. Accordingly, a total of 233 paraquat-poisoned patients were enrolled in this study. The total survival rate at discharge was 41.6% (97 of 233). Among the 233 patients, 40 patients were negative on the urine sodium dithionite test. All patients with negative urine sodium dithionite tests were also negative on the plasma sodium dithionite tests, and all patients survived, resulting in a 100% negative predictive value for the prediction of death. Among the 193 remaining patients with positive urine sodium dithionite tests, the plasma sodium dithionite tests were negative in 90 patients, equivocal in 12 patients, and positive in 91 patients. In the 3 subsets of patients, 51 of 90, 6 of 12, and none of 91 patients survived. Accordingly, the plasma sodium dithionite test had a 100% positive predictive value for the prediction of death. In 12 patients with an equivocal plasma sodium dithionite test, 6 patients in whom treatment could be started within 4 hours after ingestion survived. However, none of the remaining 6 patients with equivocal results, who presented to the hospital ⬎4 hours after ingestion, survived. Table 1 lists the principal characteristics of the patients and comparisons according to urine and plasma sodium dithionite tests. Because no patient with a positive plasma sodium dithionite test survived, the additional therapeutic benefits of the cyclophosphamide pulse and/or CVVH were evaluated retrospectively only in the subgroup of patients with negative or equivocal plasma sodium dithionite tests (Table 2). The analysis revealed no significant difference in mortality between the treatment groups. Volume 338, Number 5, November 2009
Plasma Dithionite Test in Paraquat Poisoning
TABLE 1. Comparisons according to urine (U)/plasma (P) dithionite test results Urine (U)/plasma (P) dithionite test result Variables
(U)Negative/(P)Negative (n ⴝ 40)
(U)Positive/(P)Negative (n ⴝ 90)
(U)Positive/(P)Equivocal (n ⴝ 12)
(U)Positive/(P)Positive (n ⴝ 91)
P
40.6 ⫾ 12.3 24 (60.0) 0.6 ⫾ 0.6 5.2 ⫾ 2.1 15 (37.5) 25 (62.5) 72 ⫾ 19
46.0 ⫾ 15.8 57 (63.3) 1.9 ⫾ 1.0 5.9 ⫾ 2.2 35 (38.9) 51 (56.7) 112 ⫾ 75
48.3 ⫾ 14.7 6 (50.0) 2.1 ⫾ 0.8 5.3 ⫾ 3.3 3 (25.0) 6 (50.0) 141 ⫾ 76
46.3 ⫾ 16.5 51 (56.0) 2.6 ⫾ 0.6 5.7 ⫾ 2.9 37 (40.7) 55 (60.4) 133 ⫾ 107
0.203 0.695 ⬍0.001 0.480 0.772 0.830 0.002
80 ⫾ 16
227 ⫾ 221
293 ⫾ 223
306 ⫾ 165
⬍0.001
2 (5.0) 33.8 ⫾ 28.6 27.8 ⫾ 19.3 10.4 ⫾ 5.5 1 (2.5) 0 (0)
40 (44.4) 48.0 ⫾ 52.1 31.2 ⫾ 27.1 13.0 ⫾ 6.8 48 (53.3) 39 (43.3)
7 (58.3) 69.3 ⫾ 54.5 54.7 ⫾ 32.2 17.3 ⫾ 6.8 6 (50.0) 6 (50.0)
74 (81.3) 87.8 ⫾ 133.1 51.5 ⫾ 60.2 15.4 ⫾ 15.2 86 (94.5) 91 (100)
⬍0.001 0.007 0.003 0.062 ⬍0.001 ⬍0.001
Age (yr) Male, n (%) Ingestion amount, mouthful Time to HP (hr) CVVH treatment, n (%) CYC pulse, n (%) Initial serum creatinine (mol/L) Peak serum creatinine (mol/L) Acute renal failurea, n (%) AST (U/L) ALT (U/L) Total bilirubin (mol/L) Hypoxiab, n (%) Mortality, n (%)
Values are expressed as mean ⫾ SD unless otherwise noted. a Acute renal failure: serum creatinine ⬎133 mol/L. b Hypoxia: PaO2 ⬍60 mmHg. HP, hemoperfusion; CVVH, continuous venovenous hemofiltration; CYC, cyclophosphamide; AST, aspartate transaminase; ALT, alanine transaminase.
The reproducibility of the plasma sodium dithionite test was 98% for positive cases and 92% for negative or equivocal cases with a kappa score of 0.915.
DISCUSSION Although the plasma paraquat concentration seems to be the most useful marker of the severity of paraquat poisoning,
measurements of the plasma paraquat level are not readily available in most hospitals. Moreover, most currently available methods require experience and rigorous quality control to provide reliable results.2,4 Most other prognostic indicators also require blood gas analysis and biochemical tests, such as potassium, bicarbonate, and creatinine concentrations, which may require several hours to obtain results and have relatively
TABLE 2. Comparisons according to treatment modalities in subgroup of patients with negative or equivocal plasma sodium dithionite test Treatment modalities Variables Age (yr) Male, n (%) Ingestion amount, mouthful Time to HP (hr) Initial serum creatinine (mol/L) Peak serum creatinine (mol/L) Acute renal failurea, n (%) AST (U/L) ALT (U/L) Total bilirubin (mol/L) Hypoxiab (%) Mortality (%)
HP only (n ⴝ 35)
HP ⴙ CVVH (n ⴝ 25)
HP ⴙ CYC (n ⴝ 54)
HP ⴙ CVVH ⴙ CYC (n ⴝ 28)
P
48.4 ⫾ 13.2 24 (68.6) 1.7 ⫾ 0.9 5.3 ⫾ 2.1 108 ⫾ 88 178 ⫾ 195 10 (28.6) 41.4 ⫾ 31.1 30.2 ⫾ 19.2 12.8 ⫾ 5.6 16 (45.7) 11 (31.4)
43.7 ⫾ 13.4 15 (60.0) 1.4 ⫾ 1.1 5.3 ⫾ 1.6 89 ⫾ 48 169 ⫾ 149 9 (36.0) 57.2 ⫾ 50.4 37.6 ⫾ 31.5 10.3 ⫾ 4.1 11 (44.0) 8 (32.0)
44.2 ⫾ 14.4 33 (61.1) 1.5 ⫾ 1.1 6.0 ⫾ 2.7 96 ⫾ 47 178 ⫾ 170 19 (35.2) 40.9 ⫾ 43.9 29.9 ⫾ 27.0 13.2 ⫾ 8.2 15 (27.8) 17 (31.5)
41.6 ⫾ 18.7 15 (53.6) 1.3 ⫾ 0.8 5.7 ⫾ 2.2 119 ⫾ 86 256 ⫾ 280 11 (39.3) 49.3 ⫾ 65.2 32.9 ⫾ 27.9 13.3 ⫾ 6.0 13 (46.4) 9 (32.1)
0.315 0.682 0.296 0.419 0.456 0.292 0.834 0.490 0.657 0.277 0.218 0.969
Values are expressed as mean ⫾ SD unless otherwise noted. a Acute renal failure: serum creatinine ⬎133 mol/L. b Hypoxia: PaO2 ⬍60 mmHg. HP, hemoperfusion; CVVH, continuous venovenous hemofiltration; CYC, cyclophosphamide; AST, aspartate transaminase; ALT, alanine transaminase.
© 2009 Lippincott Williams & Wilkins
375
Koo et al
poor discriminative powers for the prediction of survival or death.2,9 –12 Therefore, a simple and accurate prognostic marker is needed to help clinicians predict the prognosis and guide initial therapy. This study showed that a plasma/urine sodium dithionite test at admission can provide clinically important information to determine the severity and predict the prognosis of paraquat poisoning. In this study, the target sensitivity for plasma paraquat detection was as high as 2.0 mg/L, which obviates complex and time-consuming paraquat extraction or plasma protein precipitation steps.4 The entire procedure was completed within 10 minutes and could be performed easily without any equipment in emergency situations. The management of paraquat poisoning is directed at removing paraquat from the gastrointestinal tract, increasing its excretion from the blood, reducing radical-induced tissue injury with antioxidants, and preventing pulmonary damage with antiinflammatory drugs.1 Lin et al8 reported that an intravenous cyclophosphamide pulse might be effective in treating patients with pulmonary fibrosis due to paraquat poisoning. In a previous study,3 it was shown that the addition of prophylactic CVVH can allow for the successful management of paraquat-poisoned patients within the time where acute multiorgan failure can lead to immediate death caused by circulatory collapse. Therefore, the combination of CVVH, which prevents early death from multiorgan failure, and pulse cyclophosphamide, which prevents late pulmonary fibrosis, might be effective therapeutic options in paraquat poisoning. However, the efficacy of those intensive treatments, which are not without risk, has not been verified using a validated prognostic system. It is important to predict the prognosis or response to aggressive treatment and advise the patients and relatives of the likely outcomes. According to the results of this study, intensive treatment, including cyclophosphamide pulse and/or CVVH, had no therapeutic effect on patients with positive plasma sodium dithionite tests. Therefore, it is reasonable to use an intensive form of treatment preferentially in the subgroup of patients with negative plasma sodium dithionite tests. However, in this study, there was also no difference in mortality according to treatment modality in the subgroup of patients with negative plasma sodium dithionite tests. This might be due to the retrospective nature of the analysis. Therefore, a prospective study will be needed to evaluate the therapeutic effects of a cyclophosphamide pulse and/or CVVH in subgroups of patients with negative plasma sodium dithionite tests. There are some limitations in this study. The test is a color read, semi-qualitative test that uses the naked eye. Ac-
cordingly, rigorous standardization of the test will be needed before its clinical applications can be accepted. In our laboratory, 2 observers interpreted the results to reduce the likelihood of subjectivity in the interpretation of plasma dithionite tests, and any disagreement between the 2 observers was reported as equivocal. Moreover, the reproducibility of the plasma sodium dithionite test was 98% for the positive cases and 92% for the negative or equivocal cases with a kappa score of 0.915, which supports reliability of this test as a prognostic marker, particularly in definitely positive cases. Another limitation is that there might be cross-reactions with compounds such as diquat and interference from a high bilirubin level. However, in this study, diquat-poisoned patients and paraquat-poisoned patients, who presented to the hospital ⬎12 hours after paraquat ingestion, were excluded. Because liver failure usually begins 1 to 3 days after paraquat ingestion, hyperbilirubinemia induced by liver failure will rarely develop within 12 hours after paraquat ingestion. In this study, only 2 patients had baseline hyperbilirubinemia (plasma bilirubin ⬎2.0 mg/dL) at admission. According to the survival curve,5,6 which relates the outcome to the plasma paraquat concentration at admission and ingestion to the sampling interval, a small portion of patients with plasma paraquat concentrations ⬎2.0 mg/L have a chance of survival if treatment is begun within 2 to 4 hours after ingestion. Indeed, Gil et al7 reported that the upper limit of the plasma paraquat level in survivors was 2.64 mg/L at 3 hours. In conformity with the above-mentioned studies, 6 survivors of 12 patients with an equivocal plasma sodium dithionite test presented to the hospital within 4 hours of ingestion. This suggests that intensive treatment can be attempted with some success in select patients with a positive plasma sodium dithionite test if they present to the hospital within 2 to 4 hours after paraquat ingestion.
CONCLUSION The plasma sodium dithionite test is a simple and reliable prognostic marker with a 100% positive predictive value for the prediction of death in paraquat-poisoned patients who present to the hospital ⬎4 hours after ingestion. Aggressive treatment, including a cyclophosphamide pulse and/or CVVH, might have no therapeutic effect on patients with definitely positive plasma sodium dithionite tests. A relatively high survival rate can be expected in patients with negative plasma sodium dithionite tests. A treatment algorithm is suggested based on the plasma and urine sodium dithionite tests (Figure 1). However, considering the semi-quantitative nature of this test performed at a
Urine sodium dithionite test
Negative
Positive (or Anuric)
Minimal poisoning : 100% survival
Plasma sodium dithionite test
: Supportive care (± Hemoperfusion) Negative or Equivocal
Positive
Mild to moderate poisoning
Severe poisoning
: 50-60% survival
: 0% survival especially in patients
: May require intensive treatment (Hemoperfusion/Hemofiltration)
FIGURE 1. Suggested therapeutic approach for paraquat-poisoned patients based on the sodium dithionite test results.
presenting to hospital > 4hrs post-ingestion : Supportive care only
(Steroid/Cyclophosphamide)
376
Volume 338, Number 5, November 2009
Plasma Dithionite Test in Paraquat Poisoning
single institution, further confirmatory studies with rigorous standardization will be needed before this test can be used as a therapeutic guideline in the decision to use or withhold aggressive therapy. Until this test has been quantified and verified, it can be used to counsel victims and/or their families regarding the prognosis in an individual case. ACKNOWLEDGMENT The authors thank the staff of the clinical laboratory of the Chuncheon Sacred Heart Hospital for their assistance. REFERENCES 1. Dinis-Oliveira RJ, Duarte JA, Sa´nchez-Navarro A, et al. Paraquat poisonings: mechanisms of lung toxicity, clinical features, and treatment. Crit Rev Toxicol 2008;38:13–71. 2. Eddleston M, Wilks MF, Buckley NA. Prospects for treatment of paraquat-induced lung fibrosis with immunosuppressive drugs and the need for better prediction of outcome: a systematic review. QJM 2003;96:809 –24. 3. Koo JR, Kim JC, Yoon JW, et al. Failure of continuous venovenous hemofiltration to prevent death in paraquat poisoning. Am J Kidney Dis 2002;39:55–9. 4. Braithwaite RA. Emergency analysis of paraquat in biological fluids. Hum Toxicol 1987;6:83– 6.
© 2009 Lippincott Williams & Wilkins
5. Proudfoot AT, Stewart MS, Levitt T, et al. Paraquat poisoning: significance of plasma-paraquat concentrations. Lancet 1979;2:330 –2. 6. Lheureux P, Ekwall B. Time-related lethal blood concentrations from acute human poisoning of chemicals (The MEMO database) Part 2: the monographs. No. 25, Paraquat 1997. Available at: http://www. cctoxconsulting.a.se/25_paraquat.pdf. Accessed October 1, 2008. 7. Gil HW, Kang MS, Yang JO, et al. Association between plasma paraquat level and outcome of paraquat poisoning in 375 paraquat poisoning patients. Clin Toxicol 2008;46:515– 8. 8. Lin JL, Leu ML, Liu YC, et al. A prospective clinical trial of pulse therapy with glucocorticoid and cyclophosphamide in moderate to severe paraquat-poisoned patients. Am J Respir Crit Care Med 1999; 159:357– 60. 9. Suzuki K, Takasu N, Arita S, et al. A new method for predicting the outcome and survival period in paraquat poisoning. Hum Toxicol 1989;8:33– 8. 10. Yamaguchi H, Sato S, Watanabe S, et al. Pre-embarkment prognostication for acute paraquat poisoning. Hum Exp Toxicol 1990;9:381– 4. 11. Ragoucy-Sengler C, Pileire B. A biological index to predict patient outcome in paraquat poisoning. Hum Exp Toxicol 1996;15:265– 8. 12. Hong S-Y, Yang D-H, Hwang K-Y. Association between laboratory parameters and outcome of paraquat poisoning. Toxicol Lett 2000;118: 53–9.
377