Tox and Hound – Fellow Friday – Bipyridyl Herbicide Poisonings: Diquat and Paraquat

A crash board review for medical toxicology fellows

Steven Curry, M.D.
University of Arizona College of Medicine – Phoenix
Department of Medical Toxicology
Banner – University Medical Center Phoenix
Phoenix, AZ

Introduction

Medical toxicology fellows must know the toxicities of diquat and paraquat, as many potential board questions arise from these topics, and it’s best to have plans of action prepared in advance for when these patients arrive for care. Both diquat and paraquat share similar mechanisms of action and, initially, produce indistinguishable signs and symptoms. But there are some important differences to be aware of.

This is meant to be a brief overview of major points for boards and initial patient care following dipyridyl herbicide oral ingestion. While our group has cared for diquat- and paraquat-poisoned patients for more than 40 years, we saw these patients much more commonly in the 1980s and early 1990s than today, at which time serious ingestions are very uncommon in Arizona. Importantly, our experience is tiny compared to physicians in some parts of the world, such as Asia, where poisonings by dipyridyl herbicides are extremely common. I sent medical toxicology fellows to Taiwan for one-month rotations in the early 1990s, and they saw 2 deaths from paraquat, alone, every 3 days.

This time I thought we would try a slightly different format to promote retention of information. We’ll begin with a brief background and then center the post around a case and branch off to relevant discussions from there.

Background

Diquat and paraquat are post-emergent and non-selective dipyridyl herbicides found as dibromide or dichloride salts, respectively. They kill most plants after absorption through leaves, but are completely inactivated when in contact with soil; they have no effect on weeds yet to emerge. The products are available in various formulations (liquids, granules) under numerous brand names, and sometimes in combinations with each other or other herbicides. Because leaves rapidly die after application, diquat is also used for defoliation prior to harvesting of some crops.

Available formulations range from those that are dilute enough for direct spraying to liquid concentrates for dilution, which generally lie in the range of about 25 to 45%. While unintentional single swallows of dilute solutions meant for spraying usually produce no major morbidity or mortality, a single swallow of a concentrate carries about an 85% mortality rate. Intentional large-volume ingestions of dilute solutions also produce severe poisoning.

Immediate binding and inactivation of these products in soil led to original use of substances like bentonite clay and fullers earth for GI decontamination.

Subsequent studies showed activated charcoal adsorbed the herbicides just as well, if not better, so charcoal remains the most commonly agent used for GI decontamination.

Illustrative Case

A 32-year-old landscaper unintentionally ingested one large swallow of what he thought to be water from a cup at work. He began to vomit and sought medical care. In the ED, 2 hours later, he immediately received 100 grams activated charcoal orally. Initial vital signs included HR 158 bpm, BP 196/118 mmHg, Resp 27, and T 97.6° F. He was immediately able to provide a urine sample. Figure 3 shows his urine after addition of sodium dithionite in potassium hydroxide about 10 minutes after arrival. The dark green color indicates high concentrations of diquat (Figure 4).

Simplistically, diquat and paraquat undergo redox cycling to generate reactive oxygen species (ROS). Figure 5 shows diquat accepting electrons from NADPH to create superoxide ions which undergo conversion to H₂O₂ non-enzymatically and by superoxide dismutase. H₂O₂ then reacts with Fe²⁺ in the Fenton reaction to generate extremely reactive hydroxyl radicals, which cause lipid peroxidation, DNA damage, and other damage that produces cell injury and death. Highly unsaturated inner mitochondrial membranes are particularly susceptible to peroxidation. Early mitochondrial damage is noted with electron microscopy in animal cells.

As an aside, how does NADPH reduce oxidized diquat? Various flavoenzymes move electrons from NADPH or NADH onto oxidized diquat. The membrane-bound enzyme, cytochrome P450 reductase (CYP450 reductase), for example, normally moves electrons from NADPH onto cytochrome P450 (CYP450), such as for xenobiotic oxidative metabolism. But this enzyme can place electrons on various substrates apart from CYP450, including cytochrome b5, some drugs (e.g., mitomycin) and diquat and paraquat. CYP450 reductase uses FAD and FMN (riboflavin-5-phosphate) in the process of shuttling electrons onto substrates. Xanthine oxidase reduces diquat in rodent studies, and inhibition by allopurinol attenuates toxicity. An NADH:quinone reductase in the external mitochondrial membrane (not to be confused with complex I in the internal membrane) is also active in animal studies. There are undoubtedly other flavoenzymes that also participate in bipyridyl herbicide reduction.

Glutathione (and catalase) reduce H₂O₂ to water, and such action, in part, reduces cellular glutathione concentrations. Of course the picture is much more complicated, and generation of peroxynitrite and other radicals is involved, but we don’t need to go any deeper in this brief review. You also may see figures in which NADP⁺ is reduced back to NADPH by G6PD in the hexose monophosphate (HMP) shunt. While the HMP shunt is the only major route of NADPH generation in erythrocytes, in other cells there are multiple pathways for NADPH synthesis. So there is no reason to believe that G6PD deficiency has a significant influence on bipyridyl herbicide toxicity apart, possibly from hemolysis, which is not typically a problem.

This man had swallowed diquat concentrate (37.3%). GI injury, kidney failure, heart failure, hepatic necrosis, rhabdomyolysis, and ARDS were expected. Some patients develop encephalopathy, transient hypertension (as in this patient), seizures, adrenal necrosis, methemoglobinemia and other complications. For those who survive the initial insult, about 18% experience brainstem (e.g., midbrain, pons, etc.) infarct/hemorrhage within the first several days.

After the dithionite test, we immediately went back to the bedside and explained to the patient imminent multiple organ system failure (MOSF), the need for pressors, lines, renal replacement therapy, and the very high mortality. He requested everything be done and clarified who his decision maker would be for when he was incapacitated.

It was now about 25 minutes after arrival, and apart from supportive care, the immediate treatment strategy we used is shown in Figure 7. Our deferoxamine infusion that follows the loading dose was reduced for expected AKI.

Indeed, hypotension, oliguria, and lethargy began within about 60 minutes and he underwent endotracheal intubation; hemodialysis was begun about 2 hours after arrival. A post-intubation/line CXR is shown in Figure 8.

The next morning EGD demonstrated moderate esophagitis and some ulceration and erythema in gastric fundus – much less than we expected based on past experience (Figures 9 & 10).

The patient remained sedated on the ventilator. Cardiogenic shock/LV dysfunction was treated with pressors and inotropes. A PA catheter was placed for following continuous cardiac output. As pulmonary infiltrates developed, an FiO₂ of 0.21 was able to be maintained until day 4, when O₂ requirements began to rise to keep saturations ≥ 85%. By day 12, severe ARDS was established (Figures 11 & 12). Bronchoscopy for severe, thick secretions and plugging was required as often as TID for a few days. Neuromuscular blockade was required intermittently.

Deep sedation without NMJ blockade was used on and off and focused on alternating between various receptors or combinations, thereof: opioids (MOR) ; propofol/benzos/barbs (GABA_A); haloperidol (D₂); and dexmedetomidine (α₂). A head CT on day 12 was obtained to rule out brainstem infarct/hemorrhage and was normal (Figure 13).

Below is a CT of another one of our diquat patients (Figure 14). This man had ingested diquat concentrate and experienced moderate GI injury with AKI requiring hemodialysis, but had never been encephalopathic or hypotensive. His lungs had remained clear. On day 5 he suddenly became unconscious while resting in bed. A midbrain infarct extended into thalamus and down into the pons. The normal scan in our current pt was a relief 12 days post-ingestion.

Transaminases peaked at ~ 700 IU/L and fell. CK peaked at ~ 2500 U/L. Heart, liver, and lungs improved over 3 weeks. Oliguria remained, and intermittent hemodialysis was commenced. He underwent tracheostomy and was moved to a medical floor. A CXR on day 28 is shown in Figure 15. Shortly thereafter the tracheostomy tube was removed. He was discharged on HD at 5 weeks. After 1 year he had completely recovered, with normal pulmonary and renal function, and was back to work.

Randomized controlled trials have not convincingly demonstrated that anything apart from supportive care changes outcome. But in vitro and animal data certainly support limiting FiO₂ to minimally acceptable saturations. In fact, raising FiO₂ worsens oxidant stress and tissue injury in animals. Some in vitro and animal studies support use of deferoxamine to limit formation of ROS, but other studies have failed to find a benefit. Additional non-human data support administering NAC, vitamin E, or other general antioxidants. Oral activated charcoal certainly limits absorption if given immediately after ingestion, but hasn’t been shown to change outcome when given after the patient arrives for care.

Thus, we limit supplemental oxygen and give oral activated charcoal, deferoxamine, and NAC to those in whom survival is thought possible. We begin hemodialysis followed by CRRT immediately in order to remove any possible diquat and because kidney injury is usually established by time of or shortly after admission. However, many studies indicate removal by this method is relatively poor or insignificant and does not increase survival. Allopurinol administration seems benign, but the GI injury typically limits PO medications initially.

At times we’ve initially mixed air with N₂ to drop FiO₂ to 15% at time of admission, keeping saturations > 85%. This has been described by others, as well. The girl, below, survived ingestion of paraquat concentrate with such therapy, and her “paraquat tongue” shows the typical delayed peak onset mucous membrane injury the day after admission (Figures 16 & 17).

Paraquat Versus Diquat

Briefly, moderate to severe poisonings by diquat and paraquat initially are clinically indistinguishable, with mucositis/GI injury and multiple organ system failure, including ARDS. Kidney failure and encephalopathy may be more common in diquat toxicity, but in moderate to severe cases the incidence appears to be about the same to us. Type I and II pneumocytes and Clara cells contain a polyamine pump that takes up polyamines with nitrogens separated by about 0.6 to 0.7 nm. This pump takes up paraquat and concentrates it in the lungs. Even patients with what appears initially to be mild to moderate paraquat poisoning very commonly develop delayed onset pulmonary injury with pulmonary fibrosis and death in the subsequent days to weeks. In diquat, the distance between nitrogens prevents it from being a substrate of the pump, explaining lack of delayed pulmonary toxicity, though initial ARDS is certainly possible. The reason why brainstem infarctions are seen with diquat and typically not with paraquat is not understood.

For diquat, then, if a patient survives the initial MOSF, including ARDS, there is concern for delayed onset brainstem infarct/hemorrhage in the first week or so, but not delayed pulmonary injury/fibrosis.

For paraquat, if a patient survives the initial MOSF, which at times may be mild and may not involve the lungs, there is great concern for delayed onset pulmonary injury and fibrosis, which is commonly fatal.

Efforts to prevent concentration of paraquat in lungs or subsequent pulmonary injury have yet to be proved successful in poisoned human beings. Immediate hemodialysis and/or charcoal hemoperfusion upon arrival has not been shown to be effective, but is used aggressively by some physicians with extensive experience in treating these patients. Radiation, corticosteroids, vitamin E, cyclophosphamide, bleomycin, ascorbic acid, and other efforts have been used, but remain of unproved clinical value.

Paraquat Nomogram

Fellows should be aware of the paraquat nomogram, based on the work of several investigators (Figure 18). In most centers in the U.S., plasma paraquat concentrations must be sent to a reference laboratory and do not return for several days.

A diquat nomogram has yet to be published. However, authors have described initial plasma diquat concentrations in survivors, and in those who die. For example, Zhou and Lu reported that only two of 24 patients with an initial plasma diquat level ≥ 3 mg/L, and one of 18 with levels ≥ 9 mg/L survived.

Urine Dithionite Test

The urine dithionite (hydrosulfite) test detects both diquat (green) & paraquat (purple). It is of particular value since it can be performed rapidly and easily, and extreme results are helpful in predicting outcomes (Figure 19).

Dithionite (Figure 20) acts as a reducing agent that, in an alkaline solution, reduces diquat and paraquat to their colorful, unionized forms.

A dark purple/black color shortly after paraquat ingestion indicates near, if not complete, 100% mortality. The goal of care mainly switches to preparing the patient and family for the likely outcome. A very light blue or colorless result shortly after ingestion indicates the patient will survive.

Interpretation of the test for diquat is not as established. In our experience, dark green has always been associated with kidney failure along with GI, liver and heart injury, ARDS, and sometimes, brainstem infarct/hemorrhage. A colorless to very light yellow-green soon after ingestion result portends a good outcome.

When the test result is a moderate to strong purple or green color, but doesn’t appear so dark as to be ink, then there might be some chance for survival, and it is those patients in whom we focus aggressive efforts.

Below (Figure 23) is urine, before and after addition of KOH/dithionite, of a 92-year-old man who ingested 150 cc paraquat concentrate (Figure 21). One of our fellows was just beginning to add the alkaline dithionite solution in Figure 22. Given the test result and presence of metabolic acidosis, hypotension, and kidney failure on presentation within 3 hr of ingestion, he was provided comfort care and died within 24 hours.

A plasma paraquat concentration drawn the day after admission, shortly before his death, later returned at 7.9 mg/L and was off the chart (Figure 24).

What about our diquat patient we began with and who barely survived (Figure 25)?

His plasma diquat concentration within 5 hours of ingestion was also 7.9 mg/L, associated with a 5% survival rate in the series by Zhou and Liu. We limited FiO₂, gave activated charcoal, deferoxamine, and NAC, and rapidly began hemodialysis with the intent to increase his chance of survival, but we can’t know that it affected the successful outcome.

If you are at a treatment center that may receive patients with diquat or paraquat toxicity, you should consider having a dithionite test procedure in place and a plan of action to treat these patients. We maintain our own test kit (Figure 26) and train fellows on it every year (along with learning to smell HCN and phosphine).

Very briefly, our kit contains a stock solution of 0.1 M KOH and small plastic containers for individual use, each containing pre-weighed amounts of sodium dithionite, pipettes, and test tubes. Our kit also contains standards of diquat and paraquat for creating positive controls. Most authors describe a positive color change to first be noticeable for diquat and paraquat at about 1 mg/L. Beware that the alkaline sodium dithionite solution will undergo oxidation and then will no longer be active. Thus, a new solution should be prepared if you choose to perform daily testing during treatment. And regardless of results of your screening test, it is wise to send blood and/or urine for confirmatory testing.

Various permutations of the dithionite kit have been described. Vohra and colleagues described a version which utilized sodium bicarbonate and dithionite. Below they show, left to right, decreasing concentrations of diquat in urine (tube 1 is control), ranging from 730 mg/L in tube 2 down to 3.65 mg/L in tube 10 and 0.73 mg/L in tube 11 (Figure 27).

Summary

Entire books have been written on bipyridyl herbicide toxicity, and we haven’t even touched on kinetics, poisoning by dermal contact, poisoning in pregnancy, fingernail changes, eye injuries, and other topics. But the most important points to remember for board exams include the following:

1. Single swallows of diquat or paraquat concentrates carry very high mortalities.
2. Administration of activated charcoal as soon as possible is indicated.
3. Toxicity results from generation of reactive oxygen species via redox cycling of the herbicide.
4. Iron is involved in superoxide formation, and declines in cellular glutathione stores may contribute to toxicity.
5. Oxygen supplementation worsens outcomes in animals, so O₂ is provided only to maintain adequate hemoglobin saturations.
6. Delayed GI corrosive injury is expected.
7. Immediate hemodialysis or hemoperfusion is typically performed when survival is thought possible, but has yet to be proved to contribute to clinically important herbicide removal.
8. No treatments/antidotes are known to change outcome in humans.
9. Paraquat is concentrated in pulmonary cells by a polyamine uptake pump. If initial organ failure from paraquat is survived, patients with moderate to severe toxicity will usually develop delayed-onset pulmonary injury/fibrosis and death.
10. Diquat is not concentrated in pulmonary tissue because nitrogens are too close together. But if initial organ failure is survived, about 18% of those with moderate to severe toxicity go on to suffer a brainstem infarction/hemorrhage.
11. The paraquat nomogram can be used to estimate probability of survival after paraquat ingestion.
12. The urine dithionite (hydrosulfite) test can help with prognosis and guiding therapy. Paraquat = purple; Diquat = green.

Postscript

Here are some questions to research and discuss with faculty.

1. Sodium dithionite has also been used in the laboratory in the assessment of patients with iron poisoning. How is it used in this setting and what is the rationale for its use?
2. Talbot reported that paraquat ingestions in pregnancy resulted in high rates of fetal death, and that cord blood concentrations could be several times higher than those in maternal blood. Can you name three other toxic substances that produce higher fetal than maternal blood concentrations?
3. The reason for midbrain and pontine infarctions following diquat poisoning remains undetermined. However, the red nucleus is prominent in the midbrain. Why is this nucleus red and theoretically why might this result in increased oxidant damage compared to other brain regions following diquat ingestion?

Selected references

1. Dinis-Oliveira RJ, Duarte JA, Sánchez-Navarro A, Remião F, Bastos ML, Carvalho F. Paraquat poisonings: mechanisms of lung toxicity, clinical features, and treatment. Crit Rev Toxicol. 2008;38(1):13-71. doi: 10.1080/10408440701669959. PMID: 18161502.
2. Eddleston, M. Paraquat and Diquat. In: Brent J, et al. Critical Care Toxicology. Springer, Cham. 2017, pp. 1855-1874. https://doi.org/10.1007/978-3-319-17900-1_10.
3. Environmental Health Criteria 39. Paraquat and Diquat. World Health Organization, Geneva, 1984. https://apps.who.int/iris/bitstream/handle/10665/37301/9241540994-eng.pdf.
4. Gaudreault P, Friedman PA, Lovejoy FH Jr. Efficacy of activated charcoal and magnesium citrate in the treatment of oral paraquat intoxication. Ann Emerg Med. 1985 Feb;14(2):123-5. doi: 10.1016/s0196-0644(85)81072-6. PMID: 3970396.
5. Gil HW, Hong JR, Jang SH, Hong SY. Diagnostic and therapeutic approach for acute paraquat intoxication. J Korean Med Sci. 2014 Nov;29(11):1441-9. doi: 10.3346/jkms.2014.29.11.1441. Epub 2014 Nov 4. PMID: 25408572; PMCID: PMC4234908.
6. Hart TB, Nevitt A, Whitehead A. A new statistical approach to the prognostic significance of plasma paraquat concentrations. Lancet. 1984 Nov 24;2(8413):1222-3. doi: 10.1016/s0140-6736(84)92784-3. PMID: 6150271.
7. Huh JW, Jegal Y, Hong S-B, Oh YM, Shim TS, Lim C-M, et al.: Efficacy of deferoxamine on paraquat poisoning. Tuberc Respir Dis (Seoul) 2007;62(2):113-118. DOI: https://doi.org/10.4046/trd.2007.62.2.113 https://synapse.koreamed.org/articles/1001052.
8. Jones AL, Elton R, Flanagan R. Multiple logistic regression analysis of plasma paraquat concentrations as a predictor of outcome in 375 cases of paraquat poisoning. QJM. 1999 Oct;92(10):573-8. doi: 10.1093/qjmed/92.10.573. PMID: 10627878.
9. Jones GM, Vale JA. Mechanisms of toxicity, clinical features, and management of diquat poisoning: a review. J Toxicol Clin Toxicol. 2000;38(2):123-8. doi: 10.1081/clt-100100926. PMID: 10778908.
10. Jyothsna P, Vinapamula KS: Urine sodium dithionite test: a useful clinical test for paraquat poisoning. J Clin Sci Res 2020;9:184-185.
11. Li LR, Sydenham E, Chaudhary B, You C. Glucocorticoid with cyclophosphamide for paraquat-induced lung fibrosis. Cochrane Database Syst Rev. 2012 Jul 11;(7):CD008084. doi: 10.1002/14651858.CD008084.pub3. Update in: Cochrane Database Syst Rev. 2014;(8):CD008084. PMID: 22786512.
12. Lin G, Long J, Luo Y, Wang Y, Zewu Q. Continuous venovenous hemofiltration in the management of paraquat poisoning: A meta-analysis of randomized controlled trials. Medicine (Baltimore). 2017 May;96(20):e6875. doi: 10.1097/MD.0000000000006875. PMID: 28514303; PMCID: PMC5440140.
13. Magalhães N, Carvalho F, Dinis-Oliveira RJ. Human and experimental toxicology of diquat poisoning: Toxicokinetics, mechanisms of toxicity, clinical features, and treatment. Hum Exp Toxicol. 2018 Nov;37(11):1131-1160. doi: 10.1177/0960327118765330. Epub 2018 Mar 23. PMID: 29569487.
14. Price LA, Newman KJ, Clague AE, Wilson PR, Wenck DJ. Paraquat and diquat interference in the analysis of creatinine by the Jaffé reaction. Pathology. 1995 Apr;27(2):154-6. doi: 10.1080/00313029500169772. PMID: 7567143.
15. Proudfoot AT, Stewart MS, Levitt T, Widdop B. Paraquat poisoning: significance of plasma-paraquat concentrations. Lancet. 1979 Aug 18;2(8138):330-2. doi: 10.1016/s0140-6736(79)90345-3. PMID: 89392.
16. Senarathna L, Eddleston M, Wilks MF, Woollen BH, Tomenson JA, Roberts DM, Buckley NA. Prediction of outcome after paraquat poisoning by measurement of the plasma paraquat concentration. QJM. 2009 Apr;102(4):251-9. doi: 10.1093/qjmed/hcp006. Epub 2009 Feb 19. PMID: 19228776; PMCID: PMC2659600.
17. Scherrmann JM, Houze P, Bismuth C, Bourdon R. Prognostic value of plasma and urine paraquat concentration. Hum Toxicol. 1987 Jan;6(1):91-3. doi: 10.1177/096032718700600116. PMID: 3817835.
18. Shimada H, Hirai K, Simamura E, Pan J. Mitochondrial NADH-quinone oxidoreductase of the outer membrane is responsible for paraquat cytotoxicity in rat livers. Arch Biochem Biophys. 1998 Mar 1;351(1):75-81. doi: 10.1006/abbi.1997.0557. PMID: 9500851.
19. Talbot A: Paraquat poisoning. Journal of Japan Society for Clinical Anesthesia 1988;8(2):93-117. https://www.jstage.jst.go.jp/article/jjsca1981/8/2/8_2_93/_article/-char/en.
20. Talbot AR, Fu CC, Hsieh MF. Paraquat intoxication during pregnancy: a report of 9 cases. Vet Hum Toxicol. 1988 Feb;30(1):12-7. Erratum in: Vet Hum Toxicol 1988 Jun;30(3):245. PMID: 3354175.
21. Tanen DA, Curry SC, Laney RF. Renal failure and corrosive airway and gastrointestinal injury after ingestion of diluted diquat solution. Ann Emerg Med. 1999 Oct;34(4 Pt 1):542-5. doi: 10.1016/s0196-0644(99)80059-6. Erratum in: Ann Emerg Med 2000 Feb;35(2):199. PMID: 10499956.
22. van Asbeck BS, Hillen FC, Boonen HC, de Jong Y, Dormans JA, van der Wal NA, Marx JJ, Sangster B. Continuous intravenous infusion of deferoxamine reduces mortality by paraquat in vitamin E-deficient rats. Am Rev Respir Dis. 1989 Mar;139(3):769-73. doi: 10.1164/ajrccm/139.3.769. PMID: 2647008.
23. Vohra R, Salazar A, Cantrell FL, Fernando R, Clark RF. The poison pen: bedside diagnosis of urinary diquat. J Med Toxicol. 2010 Mar;6(1):35-6. doi: 10.1007/s13181-010-0033-6. PMID: 20229151; PMCID: PMC2861174.
24. Waintrub ML, Terada LS, Beehler CJ, Anderson BO, Leff JA, Repine JE. Xanthine oxidase is increased and contributes to paraquat-induced acute lung injury. J Appl Physiol (1985). 1990 Apr;68(4):1755-7. doi: 10.1152/jappl.1990.68.4.1755. PMID: 2347813.
25. Wang X, Wang X, Zhu Y, Chen X. ADME/T-based strategies for paraquat detoxification: Transporters and enzymes. Environ Pollut. 2021 Dec 15;291:118137. doi: 10.1016/j.envpol.2021.118137. Epub 2021 Sep 9. PMID: 34536650.
26. Wegener T, Sandhagen B, Chan KW, Saldeen T. N-acetylcysteine in paraquat toxicity: toxicological and histological evaluation in rats. Ups J Med Sci. 1988;93(1):81-9. doi: 10.1517/03009734000000041. PMID: 3376355.
27. Zhou JN, Lu YQ. Lethal diquat poisoning manifests as acute central nervous system injury and circulatory failure: A retrospective cohort study of 50 cases. EClinicalMedicine. 2022 Aug 11;52:101609. doi: 10.1016/j.eclinm.2022.101609. PMID: 35990582; PMCID: PMC9386369.