A Bitter Tale
http://www.100md.com
《新英格兰医药杂志》
In this Journal feature, information about a real patient is presented in stages (boldface type) to an expert clinician, who responds to the information, sharing his or her reasoning with the reader (regular type). The authors' commentary follows.
A 53-year-old woman presented to an outpatient urgent care clinic with persistent nausea and vomiting. On the previous evening, she had had an acute onset of nausea that was followed by vomiting and light-headedness. The vomiting had occurred approximately every hour for 18 hours. She had mild discomfort in the chest and abdomen but reported no headache, fever, chills, diarrhea, shortness of breath, or diaphoresis.
These symptoms are nonspecific. The most likely cause is gastroenteritis or gastritis. Other gastrointestinal disorders — including hepatitis, cholecystitis, and pancreatitis — are possible, especially if she is febrile. The persistence of nausea and vomiting also raises the possibility of gastrointestinal obstruction. Knowledge of the quantity and quality of the emesis might help in differentiating the level, if obstruction is indeed the cause. A central nervous system process may cause nausea and vomiting; however, the patient did not have a headache or report other neurologic symptoms. I am concerned about an atypical presentation of an acute coronary syndrome. Women are more likely than men to present without classic chest pain; nausea and abdominal discomfort may be the primary presenting symptoms.
The patient had a history of the irritable bowel syndrome, attention-deficit disorder, hypercholesterolemia, and allergic rhinitis. Her medications included atorvastatin and methylphenidate. She had gone through menopause one year earlier and had been taking estrogen and medroxyprogesterone since that time. A brother had coronary artery disease with an onset before the age of 50 years. The patient had stopped smoking at the age of 40. She reported drinking a glass of wine occasionally, with no recent increase in alcohol consumption, and exercising for three hours per week.
A prescription medication, over-the-counter drug, herb, or supplement can induce gastritis or acute hepatitis or have a neurotoxic effect that might cause her symptoms. Of her medications, statins are associated with a mild increase in the aminotransferase level, but they very rarely induce symptomatic hepatitis. The patient should be asked specifically about the use of aspirin and nonsteroidal antiinflammatory drugs, which have gastric toxicity, and acetaminophen, in view of its hepatic toxicity. The patient's postmenopausal status, her use of hormone-replacement medications for the previous year, and her brother's premature coronary disease add to my concern about an acute ischemic event.
On physical examination, the patient had a heart rate of 36 beats per minute, with a regular rhythm, and a supine blood pressure of 110/60 mm Hg. When sitting up, she had a heart rate of 38 beats per minute and a blood pressure of 90/50 mm Hg. She was afebrile and had an oxygen saturation of 98 percent while breathing room air. She had minimal tenderness in the right upper quadrant of the abdomen without guarding or rebound. The cardiac examination revealed no murmurs, gallops, or rubs. The patient's lungs were clear on auscultation. Her extremities were warm, with intact pulses and no edema. She had no impairment of cognition or memory and no focal neurologic abnormalities.
With the patient's history of vomiting and orthostasis, I am surprised she does not have tachycardia. Perhaps she has a gastrointestinal illness with an exaggerated vagal response. However, this would be an unusual cause of persistent bradycardia. I would wonder about an underlying conduction defect that predisposed her to an exaggerated vagal response. Although she engages in regular physical activity, it is unlikely that the amount of exercise she reports could account for this degree of bradycardia. Hypothyroidism might cause a sinus bradycardia, but even with increased vagal tone, her heart rate is extremely slow. It is possible that she has the sick sinus syndrome, but she is rather young for this disorder.
The patient might have an underlying infiltrative cardiomyopathy with a conduction defect, such as Lyme disease, sarcoid, or amyloidosis. Undiagnosed congenital heart disease or muscular dystrophy with heart block would be expected to present at a younger age.
I am still very concerned about an acute coronary syndrome, particularly an inferior-wall myocardial infarction with heart block. If this is not an exaggerated vagal response or active ischemia, then ingestion of a drug (such as a beta-blocker, a calcium-channel blocker, or digoxin) or other toxin that causes bradycardia needs to be considered, although the patient reports no such history. At this time, I would order an electrocardiogram, establish intravenous access, give her an aspirin, and call for an ambulance to transport her to the hospital.
The patient was transferred to the emergency department. On arrival, she continued to have bradycardia, with a blood pressure of 100/50 mm Hg. An electrocardiogram revealed a marked sinus bradycardia at a rate of 36 beats per minute with nonspecific ST–T wave abnormalities (Figure 1). Her initial creatine kinase level was 247 U per liter (normal range, 27 to 218) with an MB fraction of 5.8 ng per milliliter (normal range, 0.0 to 5.0); troponin was 0.0 ng per milliliter. The serum potassium level was 5.2 mmol per liter, and the magnesium level was 1.6 mg per deciliter (0.7 mmol per liter); other electrolyte values were normal, as was the complete blood count. The serum urea nitrogen level was 12 mg per deciliter (4.3 mmol per liter), and the creatinine level was 0.9 mg per deciliter (79.6 μmol per liter). The calcium level was 8.6 mg per deciliter (2.2 mmol per liter), with a normal albumin level. The results of liver-function tests and measurements of amylase and lipase were within normal limits. Toxicologic screening of serum and urine was performed. The findings on a chest radiograph and an ultrasound of the right upper quadrant were normal. The patient was admitted to the hospital.
Figure 1. Electrocardiogram from the Patient.
The 12-lead electrocardiogram shows sinus bradycardia and diffuse, "scooping" ST–T wave abnormalities.
The electrocardiogram shows profound sinus bradycardia, diffuse ST-segment depressions in a "scooping" or "coving" pattern, and the presence of U waves. There are no Q waves to suggest prior myocardial infarction, and there is normal R-wave progression across the precordial leads.
The diffuse nature of the ST-segment depressions suggests a drug or electrolyte imbalance rather than ischemia. Also, the cardiac enzyme measurements obtained 12 to 24 hours after the onset of the patient's symptoms are somewhat reassuring. Since she is symptomatic, serial electrocardiography, measurement of cardiac enzyme levels, and cardiac monitoring are still appropriate, but attention should be focused on diagnoses other than cardiac ischemia.
I am surprised that the serum potassium level is slightly elevated in the setting of vomiting and normal renal function. However, the presence of U waves suggests that either intracellular potassium levels may be reduced or the initial laboratory value was an error. U waves can also be a toxic effect of medication (e.g., quinidine). The sinus bradycardia in conjunction with the concave pattern of the ST depressions also raises the suspicion of a toxic effect of digoxin; however, other drugs can have this effect as well.
The patient continued to have a heart rate in the 30s, with persistent nausea and occasional vomiting. She was given 0.5 mg of atropine intravenously, which was followed by a prompt increase in the heart rate to 70 beats per minute. Additional details of her history confirmed that she had not taken any unusual medications and that her most recent refills of medications did not appear to differ from her usual pills. She reported no recent travel, hiking, or tick bites. She said that she had consumed a salad with dandelion leaves from her partner's inner-city garden two days before admission and again one day before admission.
With this additional information, a toxic ingestion seems most likely. Eating dandelion leaves is relatively safe, especially when they are eaten as part of a salad. The two most commonly reported side effects of ingesting dandelions are diuresis and an elevation in the blood glucose level.
What else might the patient have consumed or been exposed to that could cause this clinical picture? Her nausea, abdominal discomfort, and bradycardia might indicate organophosphate poisoning. However, the onset of symptoms from excessive acetylcholine activity can occur within 4 hours after pesticide exposure and routinely occurs within 12 hours, whereas this patient's symptoms started more than a day after she first ate the salad. Also, she did not have some of the other classic symptoms of organophosphate poisoning, such as headache, abdominal cramping, diarrhea, blurred vision with small pupils, excessive sweating, and increased salivation.
Since her bradycardia responded to a small dose of atropine, there is no urgent indication for placement of a temporary pacemaker. If her hemodynamic status became compromised, advanced heart block developed, or she did not have a response to atropine, she then would be a candidate for a pacemaker.
The bradycardia and nonspecific electrocardiographic changes persisted for two days after admission. On the night of the second day, she had a syncopal episode with a five-second pause on telemetry after having a bowel movement (Figure 2). Serial cardiac enzyme measurements remained normal. Repeated tests of potassium and magnesium levels were normal. Toxicologic screening was negative. The level of thyrotropin was normal, and a serologic test for Lyme disease was negative. An echocardiogram showed normal left and right ventricular size and function, without wall-motion abnormalities and with structurally normal valves. An exercise stress test with perfusion imaging revealed no evidence of ischemia. Although the patient said that she had not taken digoxin, the persistent electrocardiographic findings and the absence of a clear cause of her condition prompted a test for serum digoxin on the third hospital day, which showed a level of 1.3 ng per milliliter.
Figure 2. Rhythm Strip Obtained during Telemetric Monitoring on the Second Night of the Patient's Hospitalization.
The single-lead rhythm strip shows a five-second sinus pause with ventricular escape beats.
Why does the patient have digoxin in her serum? Possibilities include surreptitious digoxin use by the patient, poisoning by someone else, or accidental ingestion of a cardiac glycoside from consumption of plants containing the compound. Other possibilities include laboratory error and a false positive test result due to the presence of endogenous digoxin-like immunoreactive factors, as has been described, for example, in patients with renal or hepatic disease or in those taking aldosterone inhibitors, such as spironolactone. However, the symptoms and signs that are consistent with digoxin toxicity argue strongly against a false positive test result.
Given the marked bradycardia and electrocardiographic changes with a serum digoxin concentration of only 1.3 ng per milliliter and normal renal function, other sources of glycoside-containing substances should be explored, including plants and herbal medications. Such glycosides may only partly cross-react or may not be detected at all in a standard digoxin serum assay. The serum digoxin level is only a rough indication of possible toxicity. A serum digitoxin level may correlate more closely with symptoms and cardiac manifestations of toxicity.
The patient and her partner reaffirmed that the patient had not taken digoxin. However, the patient's partner reported that he was growing foxglove in his garden. The patient described where she had picked dandelion leaves for her salad, and he confirmed that this was the location where he was growing foxglove. The digitoxin level, measured on the third hospital day, was 43 ng per milliliter (therapeutic range, 10 to 32).
Digitoxin is the principal active agent in the foxglove leaf. Therefore, the patient's digitoxin level is consistent with recent ingestion of foxglove. Digitoxin differs from digoxin in that digitoxin has a much higher gastrointestinal absorption, has a longer half-life (four to six days rather than two), is more protein-bound, and is cleared less by the kidneys and metabolized more in the liver. Because the effects of digitoxin are likely to persist longer than those of digoxin, the patient should be monitored in the hospital until her symptoms and bradycardia resolve.
Should the patient be treated with digoxin-specific–antibody Fab fragments? The generally accepted indications for this therapy are persistent hyperkalemia, life-threatening ventricular or supraventricular arrhythmias, hemodynamically significant bradycardia, high-degree heart block that is unresponsive to atropine, and cardiac arrest.
The patient's nausea and light-headedness began to improve, but her heart rate remained slow, at an average of 40 beats per minute. By the seventh hospital day, her heart rate had increased to the high 40s. The patient's partner brought in a sample of the foxglove, and the patient immediately recognized it as the plant she had eaten. Analysis of the plant confirmed the presence of digitoxin. The patient was well enough to go home on the ninth hospital day, with a resting heart rate in the 50s.
Commentary
Accidental digitalis toxicity associated with consumption of plants that contain cardiac glycosides is rare.1 Although foxglove is indigenous to temperate climatic zones, few people consume it because of its bitter taste.2 Exposure is more common among infants and children under six years of age than among adults.1 Ingestion may also occur from consumption of contaminated field water near places where these plants grow, from homemade herbal preparations, and from homegrown gardens, as occurred in this case.
Toxicity due to consumption of leaves from the foxglove plant (Digitalis purpurea) produces clinical findings that are similar to those associated with an overdose of digoxin. Although some patients may present with typical side effects, such as gastrointestinal symptoms (nausea, vomiting, anorexia, or diarrhea), others may have central nervous system effects (fatigue, confusion, insomnia, or psychosis), visual effects (seeing yellow "halos" around lights, blurred or double vision, or photophobia), and cardiac effects (palpitations, light-headedness, or chest pain). Such diverse manifestations can mislead the clinician and delay the diagnosis, as occurred with this patient. It should be noted that symptoms associated with digitalis toxicity do not necessarily correspond to the serum digoxin concentration, and a person may have toxic effects at a level that is considered to be normal or therapeutic.2
In general, plant-derived cardiac glycosides have properties that affect the myocardium in a manner that is similar to that of digoxin. Cardiac glycosides enhance cardiac inotropy by inhibiting the cellular membrane Na+/K+–ATPase and ultimately increasing intracellular calcium within cardiomyocytes.3,4 In addition, cardiac glycosides promote parasympathetic activity, thereby slowing the basal resting heart rate or the ventricular response to supraventricular tachycardia. This patient's syncopal episode probably occurred because of the higher vagal state associated with having a bowel movement, coupled with the increased parasympathetic activity from foxglove-derived digitalis toxicity.
In addition to foxglove (Figure 3), several other plant and herbal sources of cardiac glycosides may be detected by serum immunoassays for digoxin or digitoxin (as in this case), including woolly foxglove (D. lanata), ornamental oleander (Nerium oleander), yellow oleander (Thevetia peruviana), squill or sea onion (Uriginea maritima), lily of the valley (Convallaria majalis), and ouabain (Strophanthus gratus).2 Another source of cardiac glycosides is venom extracted from skin glands in certain species of toads (Bufo marinus and B. alvarius). This compound has turned up in some aphrodisiacs and Chinese medications (e.g., chan su).5,6 Ingestion may cause symptoms and clinical findings similar to those of digitalis overdose, and deaths have been reported.6
Figure 3. Foxglove Plant.
Patients with cardiac-glycoside toxicity from plant or herbal sources, like those with pharmaceutical digitalis toxicity, can be treated initially with activated charcoal. If toxicity is life-threatening, administration of digoxin-specific–antibody Fab fragments should be considered for rapid reversal of the cardiac complications. Because of the large volume of distribution of cardiac glycosides, dialysis is ineffective for treating this type of digitalis toxicity.2 Early consultation with a medical toxicologist or poison control center may help the clinician to identify the toxic source and guide decisions about treatment.
Could the diagnosis have been made sooner in this case? Several clues to the diagnosis were not initially recognized. The potassium level was elevated, which is unusual in the setting of nausea and vomiting. In addition, the combination of persistent bradycardia and the patient's other electrocardiographic findings should have raised the suspicion of digitalis intoxication. Had the diagnosis been established earlier, the administration of digoxin-specific–antibody Fab fragments might have been useful. In contrast to the well-established efficacy of Fab fragments for treating pharmaceutical digitalis toxicity,2,3,7,8 there is less experience with Fab fragments for treating toxicity associated with plant-containing cardiac glycosides.9,10 Nevertheless, Fab treatment is generally benign and might have reversed the patient's symptoms and shortened her hospital course.
We are indebted to Joseph L. Dorsey, M.D., and Michael E. Hurwitz, M.D., Ph.D., for their critical review of the manuscript.
Source Information
From the Department of Medicine, Harvard Vanguard Medical Associates and Brigham and Women's Hospital (L.S.N., H.E.L.); and the Division of Cardiovascular Medicine, Brigham and Women's Hospital (M.W.F.) — both in Boston.
Address reprint requests to Dr. Newman at Brigham and Women's Hospital, 75 Francis St., Boston, MA 02115, or at lnewman1@partners.org.
References
Watson WA, Litovitz TL, Rodgers GC Jr, et al. 2002 Annual report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med 2002;21:353-421.
Hoffman BF, Bigger TJ. Digitalis and allied cardiac glycosides. In: Gilman AG, Rall TW, Nies AS, Taylor P, eds. Goodman and Gilman's the pharmacological basis of therapeutics. 8th ed. New York: Pergamon Press, 1990:814-39.
Smith TW, Haber E. Digitalis. N Engl J Med 1973;289:1125-1129.
Hauptman PJ, Kelly RA. Digitalis. Circulation 1999;99:1265-1270.
Kwan T, Paiusco AD, Kohl L. Digitalis toxicity caused by toad venom. Chest 1992;102:949-950.
Gowda RM, Cohen RA, Khan IA. Toad venom poisoning: resemblance to digoxin toxicity and therapeutic implications. Heart 2003;89:e14-e14.
Kelly RA, Smith TW. Recognition and management of digitalis toxicity. Am J Cardiol 1992;69:108G-118G.
Antman EM, Wenger TL, Butler VP Jr, Haber E, Smith TW. Treatment of 150 cases of life-threatening digitalis intoxication with digoxin-specific Fab antibody fragments: final report of a multicenter study. Circulation 1990;81:1744-1752.
Rich SA, Libera JM, Locke RJ. Treatment of foxglove extract poisoning with digoxin-specific Fab fragments. Ann Emerg Med 1993;22:1904-1907.
Thierry S, Blot F, Lacherade J-C, Lefort Y, Franzon P, Brun-Buisson C. Poisoning with foxglove extract: favorable evolution without Fab fragments. Intensive Care Med 2000;26:1586-1586.
Related Letters:
A Bitter Tale
Goyal S. B., Spodick D., Ritter M. M., Newman L. S., Feinberg M. W., LeWine H. E.(Lori S. Newman, M.D., Ph.)
A 53-year-old woman presented to an outpatient urgent care clinic with persistent nausea and vomiting. On the previous evening, she had had an acute onset of nausea that was followed by vomiting and light-headedness. The vomiting had occurred approximately every hour for 18 hours. She had mild discomfort in the chest and abdomen but reported no headache, fever, chills, diarrhea, shortness of breath, or diaphoresis.
These symptoms are nonspecific. The most likely cause is gastroenteritis or gastritis. Other gastrointestinal disorders — including hepatitis, cholecystitis, and pancreatitis — are possible, especially if she is febrile. The persistence of nausea and vomiting also raises the possibility of gastrointestinal obstruction. Knowledge of the quantity and quality of the emesis might help in differentiating the level, if obstruction is indeed the cause. A central nervous system process may cause nausea and vomiting; however, the patient did not have a headache or report other neurologic symptoms. I am concerned about an atypical presentation of an acute coronary syndrome. Women are more likely than men to present without classic chest pain; nausea and abdominal discomfort may be the primary presenting symptoms.
The patient had a history of the irritable bowel syndrome, attention-deficit disorder, hypercholesterolemia, and allergic rhinitis. Her medications included atorvastatin and methylphenidate. She had gone through menopause one year earlier and had been taking estrogen and medroxyprogesterone since that time. A brother had coronary artery disease with an onset before the age of 50 years. The patient had stopped smoking at the age of 40. She reported drinking a glass of wine occasionally, with no recent increase in alcohol consumption, and exercising for three hours per week.
A prescription medication, over-the-counter drug, herb, or supplement can induce gastritis or acute hepatitis or have a neurotoxic effect that might cause her symptoms. Of her medications, statins are associated with a mild increase in the aminotransferase level, but they very rarely induce symptomatic hepatitis. The patient should be asked specifically about the use of aspirin and nonsteroidal antiinflammatory drugs, which have gastric toxicity, and acetaminophen, in view of its hepatic toxicity. The patient's postmenopausal status, her use of hormone-replacement medications for the previous year, and her brother's premature coronary disease add to my concern about an acute ischemic event.
On physical examination, the patient had a heart rate of 36 beats per minute, with a regular rhythm, and a supine blood pressure of 110/60 mm Hg. When sitting up, she had a heart rate of 38 beats per minute and a blood pressure of 90/50 mm Hg. She was afebrile and had an oxygen saturation of 98 percent while breathing room air. She had minimal tenderness in the right upper quadrant of the abdomen without guarding or rebound. The cardiac examination revealed no murmurs, gallops, or rubs. The patient's lungs were clear on auscultation. Her extremities were warm, with intact pulses and no edema. She had no impairment of cognition or memory and no focal neurologic abnormalities.
With the patient's history of vomiting and orthostasis, I am surprised she does not have tachycardia. Perhaps she has a gastrointestinal illness with an exaggerated vagal response. However, this would be an unusual cause of persistent bradycardia. I would wonder about an underlying conduction defect that predisposed her to an exaggerated vagal response. Although she engages in regular physical activity, it is unlikely that the amount of exercise she reports could account for this degree of bradycardia. Hypothyroidism might cause a sinus bradycardia, but even with increased vagal tone, her heart rate is extremely slow. It is possible that she has the sick sinus syndrome, but she is rather young for this disorder.
The patient might have an underlying infiltrative cardiomyopathy with a conduction defect, such as Lyme disease, sarcoid, or amyloidosis. Undiagnosed congenital heart disease or muscular dystrophy with heart block would be expected to present at a younger age.
I am still very concerned about an acute coronary syndrome, particularly an inferior-wall myocardial infarction with heart block. If this is not an exaggerated vagal response or active ischemia, then ingestion of a drug (such as a beta-blocker, a calcium-channel blocker, or digoxin) or other toxin that causes bradycardia needs to be considered, although the patient reports no such history. At this time, I would order an electrocardiogram, establish intravenous access, give her an aspirin, and call for an ambulance to transport her to the hospital.
The patient was transferred to the emergency department. On arrival, she continued to have bradycardia, with a blood pressure of 100/50 mm Hg. An electrocardiogram revealed a marked sinus bradycardia at a rate of 36 beats per minute with nonspecific ST–T wave abnormalities (Figure 1). Her initial creatine kinase level was 247 U per liter (normal range, 27 to 218) with an MB fraction of 5.8 ng per milliliter (normal range, 0.0 to 5.0); troponin was 0.0 ng per milliliter. The serum potassium level was 5.2 mmol per liter, and the magnesium level was 1.6 mg per deciliter (0.7 mmol per liter); other electrolyte values were normal, as was the complete blood count. The serum urea nitrogen level was 12 mg per deciliter (4.3 mmol per liter), and the creatinine level was 0.9 mg per deciliter (79.6 μmol per liter). The calcium level was 8.6 mg per deciliter (2.2 mmol per liter), with a normal albumin level. The results of liver-function tests and measurements of amylase and lipase were within normal limits. Toxicologic screening of serum and urine was performed. The findings on a chest radiograph and an ultrasound of the right upper quadrant were normal. The patient was admitted to the hospital.
Figure 1. Electrocardiogram from the Patient.
The 12-lead electrocardiogram shows sinus bradycardia and diffuse, "scooping" ST–T wave abnormalities.
The electrocardiogram shows profound sinus bradycardia, diffuse ST-segment depressions in a "scooping" or "coving" pattern, and the presence of U waves. There are no Q waves to suggest prior myocardial infarction, and there is normal R-wave progression across the precordial leads.
The diffuse nature of the ST-segment depressions suggests a drug or electrolyte imbalance rather than ischemia. Also, the cardiac enzyme measurements obtained 12 to 24 hours after the onset of the patient's symptoms are somewhat reassuring. Since she is symptomatic, serial electrocardiography, measurement of cardiac enzyme levels, and cardiac monitoring are still appropriate, but attention should be focused on diagnoses other than cardiac ischemia.
I am surprised that the serum potassium level is slightly elevated in the setting of vomiting and normal renal function. However, the presence of U waves suggests that either intracellular potassium levels may be reduced or the initial laboratory value was an error. U waves can also be a toxic effect of medication (e.g., quinidine). The sinus bradycardia in conjunction with the concave pattern of the ST depressions also raises the suspicion of a toxic effect of digoxin; however, other drugs can have this effect as well.
The patient continued to have a heart rate in the 30s, with persistent nausea and occasional vomiting. She was given 0.5 mg of atropine intravenously, which was followed by a prompt increase in the heart rate to 70 beats per minute. Additional details of her history confirmed that she had not taken any unusual medications and that her most recent refills of medications did not appear to differ from her usual pills. She reported no recent travel, hiking, or tick bites. She said that she had consumed a salad with dandelion leaves from her partner's inner-city garden two days before admission and again one day before admission.
With this additional information, a toxic ingestion seems most likely. Eating dandelion leaves is relatively safe, especially when they are eaten as part of a salad. The two most commonly reported side effects of ingesting dandelions are diuresis and an elevation in the blood glucose level.
What else might the patient have consumed or been exposed to that could cause this clinical picture? Her nausea, abdominal discomfort, and bradycardia might indicate organophosphate poisoning. However, the onset of symptoms from excessive acetylcholine activity can occur within 4 hours after pesticide exposure and routinely occurs within 12 hours, whereas this patient's symptoms started more than a day after she first ate the salad. Also, she did not have some of the other classic symptoms of organophosphate poisoning, such as headache, abdominal cramping, diarrhea, blurred vision with small pupils, excessive sweating, and increased salivation.
Since her bradycardia responded to a small dose of atropine, there is no urgent indication for placement of a temporary pacemaker. If her hemodynamic status became compromised, advanced heart block developed, or she did not have a response to atropine, she then would be a candidate for a pacemaker.
The bradycardia and nonspecific electrocardiographic changes persisted for two days after admission. On the night of the second day, she had a syncopal episode with a five-second pause on telemetry after having a bowel movement (Figure 2). Serial cardiac enzyme measurements remained normal. Repeated tests of potassium and magnesium levels were normal. Toxicologic screening was negative. The level of thyrotropin was normal, and a serologic test for Lyme disease was negative. An echocardiogram showed normal left and right ventricular size and function, without wall-motion abnormalities and with structurally normal valves. An exercise stress test with perfusion imaging revealed no evidence of ischemia. Although the patient said that she had not taken digoxin, the persistent electrocardiographic findings and the absence of a clear cause of her condition prompted a test for serum digoxin on the third hospital day, which showed a level of 1.3 ng per milliliter.
Figure 2. Rhythm Strip Obtained during Telemetric Monitoring on the Second Night of the Patient's Hospitalization.
The single-lead rhythm strip shows a five-second sinus pause with ventricular escape beats.
Why does the patient have digoxin in her serum? Possibilities include surreptitious digoxin use by the patient, poisoning by someone else, or accidental ingestion of a cardiac glycoside from consumption of plants containing the compound. Other possibilities include laboratory error and a false positive test result due to the presence of endogenous digoxin-like immunoreactive factors, as has been described, for example, in patients with renal or hepatic disease or in those taking aldosterone inhibitors, such as spironolactone. However, the symptoms and signs that are consistent with digoxin toxicity argue strongly against a false positive test result.
Given the marked bradycardia and electrocardiographic changes with a serum digoxin concentration of only 1.3 ng per milliliter and normal renal function, other sources of glycoside-containing substances should be explored, including plants and herbal medications. Such glycosides may only partly cross-react or may not be detected at all in a standard digoxin serum assay. The serum digoxin level is only a rough indication of possible toxicity. A serum digitoxin level may correlate more closely with symptoms and cardiac manifestations of toxicity.
The patient and her partner reaffirmed that the patient had not taken digoxin. However, the patient's partner reported that he was growing foxglove in his garden. The patient described where she had picked dandelion leaves for her salad, and he confirmed that this was the location where he was growing foxglove. The digitoxin level, measured on the third hospital day, was 43 ng per milliliter (therapeutic range, 10 to 32).
Digitoxin is the principal active agent in the foxglove leaf. Therefore, the patient's digitoxin level is consistent with recent ingestion of foxglove. Digitoxin differs from digoxin in that digitoxin has a much higher gastrointestinal absorption, has a longer half-life (four to six days rather than two), is more protein-bound, and is cleared less by the kidneys and metabolized more in the liver. Because the effects of digitoxin are likely to persist longer than those of digoxin, the patient should be monitored in the hospital until her symptoms and bradycardia resolve.
Should the patient be treated with digoxin-specific–antibody Fab fragments? The generally accepted indications for this therapy are persistent hyperkalemia, life-threatening ventricular or supraventricular arrhythmias, hemodynamically significant bradycardia, high-degree heart block that is unresponsive to atropine, and cardiac arrest.
The patient's nausea and light-headedness began to improve, but her heart rate remained slow, at an average of 40 beats per minute. By the seventh hospital day, her heart rate had increased to the high 40s. The patient's partner brought in a sample of the foxglove, and the patient immediately recognized it as the plant she had eaten. Analysis of the plant confirmed the presence of digitoxin. The patient was well enough to go home on the ninth hospital day, with a resting heart rate in the 50s.
Commentary
Accidental digitalis toxicity associated with consumption of plants that contain cardiac glycosides is rare.1 Although foxglove is indigenous to temperate climatic zones, few people consume it because of its bitter taste.2 Exposure is more common among infants and children under six years of age than among adults.1 Ingestion may also occur from consumption of contaminated field water near places where these plants grow, from homemade herbal preparations, and from homegrown gardens, as occurred in this case.
Toxicity due to consumption of leaves from the foxglove plant (Digitalis purpurea) produces clinical findings that are similar to those associated with an overdose of digoxin. Although some patients may present with typical side effects, such as gastrointestinal symptoms (nausea, vomiting, anorexia, or diarrhea), others may have central nervous system effects (fatigue, confusion, insomnia, or psychosis), visual effects (seeing yellow "halos" around lights, blurred or double vision, or photophobia), and cardiac effects (palpitations, light-headedness, or chest pain). Such diverse manifestations can mislead the clinician and delay the diagnosis, as occurred with this patient. It should be noted that symptoms associated with digitalis toxicity do not necessarily correspond to the serum digoxin concentration, and a person may have toxic effects at a level that is considered to be normal or therapeutic.2
In general, plant-derived cardiac glycosides have properties that affect the myocardium in a manner that is similar to that of digoxin. Cardiac glycosides enhance cardiac inotropy by inhibiting the cellular membrane Na+/K+–ATPase and ultimately increasing intracellular calcium within cardiomyocytes.3,4 In addition, cardiac glycosides promote parasympathetic activity, thereby slowing the basal resting heart rate or the ventricular response to supraventricular tachycardia. This patient's syncopal episode probably occurred because of the higher vagal state associated with having a bowel movement, coupled with the increased parasympathetic activity from foxglove-derived digitalis toxicity.
In addition to foxglove (Figure 3), several other plant and herbal sources of cardiac glycosides may be detected by serum immunoassays for digoxin or digitoxin (as in this case), including woolly foxglove (D. lanata), ornamental oleander (Nerium oleander), yellow oleander (Thevetia peruviana), squill or sea onion (Uriginea maritima), lily of the valley (Convallaria majalis), and ouabain (Strophanthus gratus).2 Another source of cardiac glycosides is venom extracted from skin glands in certain species of toads (Bufo marinus and B. alvarius). This compound has turned up in some aphrodisiacs and Chinese medications (e.g., chan su).5,6 Ingestion may cause symptoms and clinical findings similar to those of digitalis overdose, and deaths have been reported.6
Figure 3. Foxglove Plant.
Patients with cardiac-glycoside toxicity from plant or herbal sources, like those with pharmaceutical digitalis toxicity, can be treated initially with activated charcoal. If toxicity is life-threatening, administration of digoxin-specific–antibody Fab fragments should be considered for rapid reversal of the cardiac complications. Because of the large volume of distribution of cardiac glycosides, dialysis is ineffective for treating this type of digitalis toxicity.2 Early consultation with a medical toxicologist or poison control center may help the clinician to identify the toxic source and guide decisions about treatment.
Could the diagnosis have been made sooner in this case? Several clues to the diagnosis were not initially recognized. The potassium level was elevated, which is unusual in the setting of nausea and vomiting. In addition, the combination of persistent bradycardia and the patient's other electrocardiographic findings should have raised the suspicion of digitalis intoxication. Had the diagnosis been established earlier, the administration of digoxin-specific–antibody Fab fragments might have been useful. In contrast to the well-established efficacy of Fab fragments for treating pharmaceutical digitalis toxicity,2,3,7,8 there is less experience with Fab fragments for treating toxicity associated with plant-containing cardiac glycosides.9,10 Nevertheless, Fab treatment is generally benign and might have reversed the patient's symptoms and shortened her hospital course.
We are indebted to Joseph L. Dorsey, M.D., and Michael E. Hurwitz, M.D., Ph.D., for their critical review of the manuscript.
Source Information
From the Department of Medicine, Harvard Vanguard Medical Associates and Brigham and Women's Hospital (L.S.N., H.E.L.); and the Division of Cardiovascular Medicine, Brigham and Women's Hospital (M.W.F.) — both in Boston.
Address reprint requests to Dr. Newman at Brigham and Women's Hospital, 75 Francis St., Boston, MA 02115, or at lnewman1@partners.org.
References
Watson WA, Litovitz TL, Rodgers GC Jr, et al. 2002 Annual report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. Am J Emerg Med 2002;21:353-421.
Hoffman BF, Bigger TJ. Digitalis and allied cardiac glycosides. In: Gilman AG, Rall TW, Nies AS, Taylor P, eds. Goodman and Gilman's the pharmacological basis of therapeutics. 8th ed. New York: Pergamon Press, 1990:814-39.
Smith TW, Haber E. Digitalis. N Engl J Med 1973;289:1125-1129.
Hauptman PJ, Kelly RA. Digitalis. Circulation 1999;99:1265-1270.
Kwan T, Paiusco AD, Kohl L. Digitalis toxicity caused by toad venom. Chest 1992;102:949-950.
Gowda RM, Cohen RA, Khan IA. Toad venom poisoning: resemblance to digoxin toxicity and therapeutic implications. Heart 2003;89:e14-e14.
Kelly RA, Smith TW. Recognition and management of digitalis toxicity. Am J Cardiol 1992;69:108G-118G.
Antman EM, Wenger TL, Butler VP Jr, Haber E, Smith TW. Treatment of 150 cases of life-threatening digitalis intoxication with digoxin-specific Fab antibody fragments: final report of a multicenter study. Circulation 1990;81:1744-1752.
Rich SA, Libera JM, Locke RJ. Treatment of foxglove extract poisoning with digoxin-specific Fab fragments. Ann Emerg Med 1993;22:1904-1907.
Thierry S, Blot F, Lacherade J-C, Lefort Y, Franzon P, Brun-Buisson C. Poisoning with foxglove extract: favorable evolution without Fab fragments. Intensive Care Med 2000;26:1586-1586.
Related Letters:
A Bitter Tale
Goyal S. B., Spodick D., Ritter M. M., Newman L. S., Feinberg M. W., LeWine H. E.(Lori S. Newman, M.D., Ph.)