ACMT Clinicopathological Conference: The Intoxicated, Poisoned Groundskeeper

Case Presentation—Dr. Keenan

Our patient is a 37-year-old male groundskeeper at a golf course, who presented to the hospital around 11 AM with a chief complaint of intoxication. His co-workers called emergency medical services because they saw him acting abnormally, drinking alcohol, and possibly drinking from another bottle at work. Shortly after arriving to the emergency department (ED), the patient became agitated and fell off the stretcher. He was then intubated to protect his airway and the ED staff, as well as to facilitate the medical workup.

At this point, very little was known about the patient. His past medical history, past surgical history, home medications, allergies, and family history were all unknown. The only aspect of the social history that was known was his occupation.

Initial vital signs were as follows: heart rate (HR), 170 beats per minute (bpm); respiratory rate, 24 breaths per minute; blood pressure (BP), 127/80 mmHg; oxygen saturation, 97% room air; and temperature (temp), 36.8 °C. Post intubation, his HR had improved to 140 bpm, BP 104/76 mmHg, temp 36.8 °C, and respiratory rate 20 breaths per minute on the ventilator while saturating 99% on 40% FiO₂.

Physical examination conducted about 1 h after intubation was generally normal. The patient’s pupils were 4 mm and reactive. Mucus membranes were appropriately moist. Cardiac auscultation revealed a regular tachycardia. Lungs were clear to auscultation. The abdomen was soft and with normal bowel sounds. Skin was normal without flushing or diaphoresis. The neurologic exam revealed a sedated patient, no focal deficits, normal reflexes, normal tone, and a lack of clonus.

Initial lab work included a complete blood count (CBC); a basic metabolic panel; a hepatic function panel; and ethanol, acetaminophen, and salicylate concentrations. This workup was remarkable for mild leukocytosis of 12.5 × 10⁹/L, sodium 140 mEq/L, potassium 3.0 mEq/L, bicarbonate (HCO₃⁻) 21 mEq/L, creatinine 0.9 mg/dL, AST 79 U/L, ALT 80 U/L, ethanol 257 mg/dL, and undetectable acetaminophen and salicylate levels. An electrocardiogram (EKG) was performed which demonstrated sinus tachycardia with normal intervals. A chest x-ray showed appropriate endotracheal tube placement and was otherwise normal. A head CT was normal.

Poison control was contacted around 4 h after arrival to the emergency department. The toxicologist on call was contacted, and they made additional recommendations for testing and continued supportive care.

Overnight, the patient’s condition worsened. He developed a fever up to 38.9 °C and was placed on a cooling blanket. He developed worsening hypotension requiring pressor support with phenylephrine. He developed a lactic acidosis with a peak lactate of 10.7 mmol/L. See Table 1 for more details of relevant laboratory and vital sign changes.

Table 1.

Progression of relevant laboratory and vital sign changes

	ED (11 am)	11 pm	3 am	7 am
HR (bpm)	140	160	70	70
BP (mmHg)	104/76	92/63 On phenylephrine	95/71 On phenylephrine	97/75 On phenylephrine
Temp (°C)	36.8	38.9 On cooling blanket	37.9 On cooling blanket	37.1 On cooling blanket
HCO3− (mEq/L)	21	19	18	Not available
Lactate (mmol/L)	Not available	10.7	5.6	4.0

Shortly after the on-call toxicologist was made aware of these overnight changes, the family of the patient was able to obtain a photo of the bottle the patient had drank from in addition to the ethanol.

Case Discussion: The Intoxicated, Poisoned Groundskeeper—Dr. Nemanich

This case opens many possibilities for speculation. We know from the initial labs that the patient was drunk. This is surprising since he presented at 11 am on a workday. What else did he take, or get exposed to? Was he trying to get high? Did he accidentally drink out of the wrong bottle? Was this a suicide attempt? There was very little information available when the patient first arrived at the emergency department. To solve this case, exploring the setting where the patient worked and was poisoned will be key.

Popular culture is rife with clichés related to intoxicated people on golf courses. Golf humor inevitably turns to references of drunken escapades gone wrong, or golfers mistakenly believing that they play better when intoxicated. Even golfers representing themselves in publications such as Golf Digest do not shy away from this association. One of Golf Digest’s most popular videos ever posted features golfers drinking steadily over 6 h and demonstrating the effects on their golf game [1]. And the magazine’s podcast Local Knowledge featured an episode discussing the pros and cons of alcohol on the golf course, stating “the game’s relationship with alcohol—beverage carts, 19th holes, member-guests that devolve into bacchanalian feasts—has always been layered” [2].

The patient at the center of our case, however, is not a golfer. The fact that he is an employee at the golf course in the role of groundskeeper makes his foray into bacchanalian indulgence and self-poisoning more surprising and interesting. The trope of the intoxicated groundskeeper also crops up frequently in popular culture. Of note, Business Insider published an article called “The 17 Jobs Where You’re Most Likely to Become an Alcoholic” and gardeners/groundskeepers were noted to be 1.5 times more likely to die of alcoholism than average [3]. One of the most obvious parallels in popular culture to our patient and his situation is the character of Carl Spackler, the groundskeeper in the 1980 sports-comedy film Caddyshack, famously played by Bill Murray. Carl is a quirky character who rolls enormous joints, drinks alcohol frequently, talks to himself, and is obsessed with a gopher-killing crusade, all while mocking the privileged golfers to whom he is expected to be deferential.

Turning back to our patient, we know from the first set of labs that his blood ethanol concentration was 257 mg/dL when he arrived. Presumably, he was drinking the alcohol intentionally. But did he get into something else, as suspected? We are ambiguously told that he “may” have drunk from “another” bottle while at work. Unfortunately, he was too altered to describe his symptoms or provide a history, so we only have the objective data from his presentation to go on.

The patient was agitated and tachycardic on presentation to the ED and subsequently became hypotensive and hyperthermic with a significant lactic acidosis. Was he using other drugs of abuse? If the activities depicted in Caddyshack bear any connection to reality, we can hypothesize that cocaine and cannabis use are expected among golfers. And for an interesting real-world example linking golf to illicit stimulants, an article from KFOR News describes a meth lab which was found in a golf course outhouse in Oklahoma in 2013. This was discovered when golf course staff entering the outhouse noticed “strange sport drink bottles with chemicals inside” [4]. Interestingly, there are other examples of meth labs being created on or adjacent to golf courses in 2013. One of which, in South Carolina, involved meth chemicals placed in weed killer containers [5]. The groundskeeper of our case certainly would have easy access to outhouses and storage sheds and weed killer containers, and it’s conceivable that he could have a lucrative side-hustle making meth. And once drunk and disinhibited, he may have decided to indulge in the supply. Or perhaps once intoxicated, he fell victim to the age-old toxicologic trap of chemicals stored in beverage containers. Having completed my residency in Seattle, I have a bias to immediately think of meth when a patient arrives appearing intoxicated with agitation and hyperthermia. But how well does meth, or do stimulants in general, fit with the clinical data?

A stimulant could explain the patient’s agitation, significant tachycardia, and hyperthermia, and could possibly cause his elevated lactate. However, his initial normal BP progressing to hypotension overnight does not fit with a sympathomimetic toxidrome. And while hyperthermia can be a component of severe sympathomimetic toxidromes, the delayed component to the hyperthermia here is atypical. Additionally, the patient’s pupils were not dilated, but 4 mm and reactive, and his skin was not diaphoretic—contrary to the classic findings for stimulant intoxication. Of course, the fact that alcohol is playing a role in the presentation and the fact that this is presumed to be a mixed ingestion are important to keep in mind. We can’t rule out that the combination of ethanol and the sedative medications initiated after intubation may have blunted some of the classic sympathomimetic findings.

An anticholinergic toxidrome is another possibility to consider. As a groundskeeper, the patient may have access to anticholinergic plants such as jimson weed which he could have made into a tea, for example. The patient’s tachycardia, agitation, and hyperthermia fit with an anticholinergic toxidrome as well as a sympathomimetic one. But his pupils are midsize and reactive where we’d expect dilated and minimally reactive. And while an anticholinergic presentation would classically involve dry mucous membranes, hypoactive bowel sounds, and flushed skin, this patient has moist mucous membranes, normal bowel sounds, and normal skin color. Overall, anticholinergic toxicity seems somewhat less likely than sympathomimetic.

Serotonin syndrome involves another constellation of symptoms to consider in comparison to this patient’s presentation. In terms of a triggering serotonergic agent, we don’t have anything specific to go on. The patient could be on home medications which we don’t know about, however, and certainly a drug of abuse could lead to serotonin syndrome. Similar to the toxidromes discussed above, the tachycardia, agitation, and hyperthermia are congruent with serotonin syndrome. But the normal BP progressing to hypotension, the normal pupils as opposed to dilated, and the normal skin exam as opposed to diaphoresis are not consistent. More specifically relevant, the neurologic exam was unremarkable, with no clonus or rigidity.

A seizure or other etiology causing abnormal muscle contraction could account for the patient’s elevated lactate, and potentially the hyperthermia. No abnormal movements were mentioned in this patient’s hospital course. He did have “agitation” which was apparently severe enough to cause him to fall off his stretcher, but no loss of consciousness or seizure-like movements were reported. However, for the sake of thoroughly exploring the differential, as well as taking into consideration the specific setting of the golf course, there is one rare toxin that warrants discussion. What causes delayed hyperthermia and lactic acidosis due to abnormal muscle contraction and might be found on a golf course? Caddyshack’s groundskeeper Carl and his gopher-killing crusade inevitably lead me to think of strychnine, since gopher bait is the one notorious modern sources of this rare poison.

The mechanism of action of strychnine is loss of inhibitory signaling in the nervous system, primarily the spinal cord, due to inhibition of glycine binding. The loss of inhibitory effect results in generalized muscle contraction which can cause the characteristic contortion associated with the poison called opisthotonus, described as looking like “awake seizures.” However, as the unique abnormal movements are the most classic presentation of this toxin, I think we must assume this would have been commented on at some point in the case if strychnine were what the patient ingested.

Taking a closer look at the patient’s abnormal labs for clues, we see that in addition to the significant lactate elevation of 10.7 mmol/L, the patient was mildly hypokalemic at 3.0 mEq/L. He also has a mildly elevated white blood cell count which is non-specific, and mild transaminase elevation. Methylxanthines can cause hypokalemia due to intracellular shift of potassium as well as increased excretion of potassium from a diuretic effect. Methylxanthines can also cause an elevated lactate due to increased metabolic activity and muscle activity. Hyperthermia could be present due to the same mechanisms, although generally only in very severe overdoses. Caffeine would be the most plausible exposure in this class of methylxanthines. How well does caffeine fit with other aspects of the patient’s presentation? The severe tachycardia is consistent with this exposure, as is the delayed hypotension, which can occur due to vasodilation from beta agonism. The patient would have had to ingest quite a large amount or a high concentration of caffeine to get these effects, but perhaps he was trying to counteract being drunk?

As we have seen, there are many exposures and syndromes that at least partially fit with this patient’s presentation. But as the famous golfer Arnold Palmer said, “timing is everything, in life and in golf.” The timing of this case is key, in the way that the hyperthermia and lactic acidosis evolved over time and were not present on the initial presentation. While the patient could have been using stimulants, anticholinergics, or methylxanthines and may have been out in the sun as well, these exposures would be expected to cause hyperthermia on initial presentation rather than in a delayed fashion.

One important class of poisons which can cause delayed hyperthermia and lactic acidosis are the electron transport chain (ETC) inhibitors or uncouplers. Of all the mechanisms we’ve explored so far, this may fit the best to explain the patient’s delayed hyperthermia and elevated lactate. The unique setting of this case on the golf course is equally important, in terms of honing in on specific exposures. It turns out that there are quite a few ETC disruptors which might plausibly be used on a golf course, in the form of herbicides, rodenticides, insecticides, or wood preservatives. A different, but related, pesticide mechanism is inhibition of the citric acid cycle. This is the mechanism of sodium monofluoroacetate, sold as Compound 1080. But while this chemical is associated with high lactic acid levels, it does not cause hyperthermia.

Dinitrophenol (DNP) is one of the most infamous uncouplers of oxidative phosphorylation and has historically been used as both a pesticide and herbicide, as well as a wood preservative. It classically causes severe hyperthermia which typically requires aggressive management to bring down. Based on the information presented in this case, this patient’s temperature came down with a cooling blanket alone, possibly aided by his post-intubation sedation. Maybe he took a small dose, but given what a strong uncoupler DNP it is, it seems somewhat unlikely that if it were responsible for the patient’s severe tachycardia and hyperthermia to 38.9 °C, the temperature would come down that easily. More importantly, DNP is not in use anymore in the USA as a pesticide or herbicide and was withdrawn from agricultural use in 1998. It can certainly still be obtained, but it is most typically sold online as an illicit diet supplement in pill form [6]. We are told in the case presentation that the patient was “possibly drinking from another bottle.” The suspected liquid form of the xenobiotic he ingested will be helpful in ruling out which ETC disruptors are less likely. For example, bromethalin is another potent ETC uncoupler used as a rodenticide, but it is sold in the form of rodent bait mimicking worms, or as chunks or flakes. It is not in common use in a liquid form. This makes bromethalin less likely.

The insecticide chlorfenapyr is an intriguing possibility to consider. It is a pyrrole insecticide which is a pro-insecticide, metabolized in the liver to the more toxic form called tralopyril, which is the specific xenobiotic that uncouples oxidative phosphorylation. It can cause neurotoxicity leading to altered mental status and hyperthermia, symptoms which fit with the presentation of our patient, but these features typically occur multiple days after the initial exposure even in large ingestions that ultimately turn out to be fatal. The timeline does not fit with this case, where the hyperthermia and elevated lactate were present within 24 h of the ingestion. Additionally, this insecticide is most commonly used in malaria-endemic regions where insects have developed resistance to first-line agents.

Another class of ETC disruptors to consider are the chlorophenols. Both tetrachlorophenol and pentachlorophenol have been used as pesticides and herbicides, and pentachlorophenol as a molluscicide and fungicide as well. Pentacholorophenol in particular is a strong uncoupler of oxidative phosphorylation. But similar to DNP, its use is heavily restricted in the USA now, and it is no longer common. Now, its primary use in the USA is as a wood preservative, primarily on utility poles [7]. It is typically not sold to consumers and would be less likely to be found in a groundskeeper’s shed. However, the golf course where this patient works and apparently gets drunk on the job may not be a paragon of modernization. I don’t think we can entirely rule out the possibility that there could be some DNP or pentachlorophenol still hanging around.

Finally, there are the chlorophenoxy herbicides, of which 2,4-Dichlorophenoxyacetic acid (2,4-D) is most commonly used. This is another inhibitor of oxidative phosphorylation, although not as potent as DNP or pentachlorophenol. Looking at the label of a commonly sold brand, it indicates that the product is specifically intended “for broadleaf weed control in: pastures and rangeland, lawns, golf courses, cemeteries, and similar ornamental turf.” So, this is something that would be very possible for our groundskeeper to have on hand. Additionally, it’s important to note that it is usually sold as a concentrate and must be diluted. While it’s somewhat implausible to imagine our patient drinking directly from an herbicide container unless he were suicidal, it becomes a classic tox mishap if it were diluted and stored in some type of generic bottle or drinking container. Especially with the patient being drunk, it could be an easy mistake to make.

The question is how well does 2,4-D fit the clinical presentation? Case reports show us that ingestions of 2,4-D often present with early tachycardia, and hyperthermia may be present as well. Additionally, hypotension and a range of neurologic effects can develop early on, including agitation. And the ETC inhibition leads to a delayed lactic acidosis [8]. Some classic findings in 2,4-D poisoning which are not mentioned in this case, however, are GI symptoms and muscle spasms. Additionally, renal injury is common. Our patient had a very slight increase in his creatinine from 0.9 to 1.2 mg/dL despite presumed fluid resuscitation but did not develop acute kidney injury.

To further explore the golf course setting given the multiple possibilities for electron transport disruptors and other toxins, I naturally had to visit a golf course to gain some additional insight and search for clues. But not just any golf course would do. There are a variety of exclusive, well-manicured golf courses in the Chicago area where I live. Not only are those harder to snoop around, but they also don’t seem to be the type of place where a Carl Spackler-like, intoxicated groundskeeper would likely be bumbling around. The Columbus Park golf course on the border of Oak Park and Chicago, on the other hand, is a different scene. It is not fenced off and anyone can walk in. Many characters besides myself were wandering around there with no intention of golfing. The grounds have an unkempt beauty to them. There’s a large pond surrounded by tall grass and trees and an old, dilapidated, formerly grand building at the center, as well as some practical-looking small buildings.

I eventually found a golf course employee who was willing to show me into the maintenance shed on the grounds. In the midst of cleaning products, grass seed, hoses, and inexplicably, a dollhouse full of rusty nuts and bolts, I found a shelf with two different types of broadleaf herbicides as well as lawn fertilizer. One of the herbicides was Triclopyr, which acts like a plant hormone and does not inhibit electron transport or result in the kind of presentation our patient experienced. The other herbicide on the shelf was 2,4-D (see Fig. 1).

Fig. 1.

Products on a shelf in the maintenance shed, as discovered by Dr. Nemanich. The shorter container in the left-foreground is labeled “2,4-D AMINE WEED KILLER”

Taking everything into consideration here, I am convinced that 2,4-D is most likely to be the mystery substance which the patient drank. The patient’s clinical presentation involving initial tachycardia and agitation, progressing to hypotension, along with the key findings of delayed hyperthermia and lactic acidosis strongly suggests an ETC uncoupler. The setting of the golf course makes an herbicide plausible. In addition, the fact that dilution is required for most preparations of 2,4-D means it is more likely to be stored in a container which inadequately identifies it, and there is more of a risk that someone would unintentionally ingest it. And finally, the fact that 2,4-D was the one ETC uncoupler that I encountered in the best re-creation of the patient’s setting I could muster…as Carl Spackler said in Caddyshack, “I got that going for me, which is nice.”

Clinical Diagnosis

2,4-dichlorophenoxy acetic acid poisoning.

Case Discussion and Outcome—Dr. Keenan

Shortly after the toxicology team was made aware of the patient’s changes overnight, the family provided a photo of the product (Fig. 2). The product contained phenoxy herbicides, of which the highest concentration was 2,4-dichlorophenoxy acetic acid (2,4-D) at 30.89%, followed by 4-chloro-2-phenoxy acetic acid (MCPA) at 8.23%. It also contained 2.77% 3,6-dichloro-2-methyoxybenzoic acid (dicamba), and the remaining 58.11% were other ingredients. The patient was demonstrating toxicity consistent with phenoxy herbicide poisoning. Therefore, treatment with alkalinization and hemodialysis was recommended and performed. The patient’s laboratory abnormalities improved, his pressor requirement was weaned, and his active cooling was stopped. He was extubated less than 48 h after presentation. Of note, he did develop rhabdomyolysis with a peak creatinine kinase of 1892 U/L.

Fig. 2.

Product ingested by patient

The phenoxy herbicides include a collection of similar molecules, including 2,4-D and MCPA. Dicamba is technically a benzoic acid herbicide but is often combined in discussion with phenoxy herbicides due to similar mechanism of action and clinical manifestations [9]. These herbicides function in agriculture by serving as analogues to the plant hormone auxin [10–12]. Auxin serves robust and diverse roles in plant physiology, with the degree of complexity perhaps best summarized by the title of a review article on the subject: “Auxin: Simply Complicated” [11]. Interestingly, the phenoxy herbicides are specific for dicot plants as opposed to monocots. As such, the herbicide can be applied to land with grass or grain crops without harming the desired plants. The binding of the phenoxy herbicides to auxin binding sites leads to multiple alterations leading to the death of the plant. For example, the auxin analogue herbicides lead to upregulation of ethylene, which leads to, among other things, abnormal leaf orientation, tissue swelling, and senescence. The binding of the phenoxy herbicides to the auxin binding site also ultimately results in upregulation of abscisic acid, which leads to a variety of negative effects, including but not limited to stomatal closure, growth inhibition, and reactive oxygen species overproduction [12]. The result of all these pathways is rapid plant death.

Fortunately, phenoxy herbicide poisonings are relatively rare poisonings in the USA, with only 1339 mentions of phenoxy herbicides in the 2021 National Poison Data System annual report out of 2,080,917 human exposures [13]. Of these 1339 mentions, only 13 were intentional, and only 4 were coded as a major outcome. There were no deaths.

In humans, the effects of phenoxy herbicides (and dicamba) are broad. Several review articles and case series describe the common effects in detail [7, 9]. Neurologically, the patients can develop agitation and/or coma. From a gastrointestinal (GI) perspective, patients can develop nausea, vomiting, and gastric irritation. From a cardiovascular perspective, patients can develop non-specific EKG changes, QTc prolongation, and hypotension. Metabolically, the phenoxy herbicides act as uncouplers—leading to a high anion gap metabolic acidosis with an elevated lactate, as well as hyperthermia. Finally, patients can develop rhabdomyolysis.

The treatment of a patient suffering from phenoxy herbicide poisoning should start with appropriate supportive care. The utility of GI decontamination is likely limited due to rapid absorption, as would be expected in most liquid ingestions. The rapid absorption was further demonstrated in a 1974 study in which healthy volunteers were administered 5 mg/kg 2,4-D [14]. This same study also demonstrated that most of the ingested 2,4-D is excreted unchanged in the urine. Given that phenoxy herbicides are weak acids, this pattern of elimination potentially lends itself to enhancement with urinary alkalinization to facilitate ion trapping. Although limited data is available, several case reports describe success with this method [15, 16]. Given the logical mechanism, relative safety of its use, and the supporting case reports, urinary alkalinization should be employed for patients suffering from phenoxy herbicide poisoning. Finally, there has been success with hemodialysis to enhance elimination, based on limited evidence [17]. It would be reasonable to perform hemodialysis in a patient severely poisoned with phenoxy herbicides, but the exact criteria are not clearly defined.

Final Diagnosis

Phenoxy herbicide poisoning.

Sources of Funding

None.

Declarations

Conflicts of Interest

None.