Suicides and therapeutic misadventures from colchicine poisoning are well documented in the literature (1–3). The toxic potential of the drug has been known since the late third century BC, according to the epigrammatist Dioscorides, who illustrated life in Egypt with word pictures (4,5). Medicinal preparations of the colchicine plant, Colchicum autumnale, were recommended for people with joint pain as early as the sixth century AD, and Benjamin Franklin is credited with introducing this therapy into the United States in 1763 for the treatment of gout. Pellsiter and Caventou first isolated the alkaloid colchicine from colchicum in 1820 (4).
Colchicine can be either injected intravenously or taken in tablet form; it is used to treat acute gouty arthritis. Phagocytosis of urate crystals is not prevented by neutrophils; however, the glycoprotein that causes joint pain and inflammation is either not produced or not released by neutrophils because of the action of colchicine. In addition, the drug has an antimitotic effect, which arrests cell division in the metaphase stage. Therefore, the cells with the highest turnover rate are affected: epithelial cells of the gastrointestinal tract and the bone marrow. About 10 hours after ingestion, the suppression of cell division is maximal (6).
The use of colchicine as a suicidal agent and its role in therapeutic misadventures has been reported. Professional journals describe several such cases (1–3). However, the use of colchicine as an intentional poisoning agent is extremely rare. A search of the world’s literature over the past 35 years did not yield any report of intentional poisoning or homicide (7–12).
Thus, it is significant that in a 6-year period, homicides caused by this drug were identified from two medical examiners’ offices. These deaths are described here, each the result of intentional poisoning in persons who were not taking the drug for therapeutic reasons.
The patient, a married 41-year-old man in reported good health, was admitted to the hospital for nausea, vomiting, profuse diarrhea, and severe abdominal pain. These symptoms he attributed to “food poisoning” from “bad” potato salad he had eaten the day before. His medical history included an admission to the same hospital approximately 1 month earlier with an acute onset of nausea and vomiting without diarrhea. The hospital evaluation at the previous admission had revealed focal esophagitis; laboratory tests had shown only elevated liver enzymes, and he was released from the hospital several days later. On the second and last admission to the hospital, the results of physical examination were unremarkable except for bloody diarrhea. Laboratory tests again showed evidence of hepatic injury and dehydration. An electrocardiogram revealed nonspecific ST abnormalities and an atrioventricular conduction delay. After 48 hours in the hospital with his wife by his side, the patient became profoundly short of breath and had a grand mal seizure, followed by cardiopulmonary arrest. Resuscitative efforts were unsuccessful. The treating physician requested an autopsy. The death did not fall under the criteria of a medical examiner’s case because the patient had been admitted to the hospital suffering from an unknown but presumed natural disease and had not died within the first 24 hours of admission. Consequently, a hospital autopsy was performed at the physician’s request.
A complete autopsy failed to show any significant abnormalities. The final diagnoses included acute cardiac arrest of undetermined causes and central lobular hepatocellular changes consistent with hypotension or toxic injury. The results of a toxicology screen, performed on serum obtained at the time of admission, were negative. The manner of death was determined to be natural, and the body was buried.
Several months later, the police were informed by a prosecuting attorney and his brother-in-law, who owned a chemical supply store, that the wife of the deceased had been in his store several months before her husband’s death to buy colchicine. At the time she purchased the entire stock of 30 g, stating to the owner that she was using the substance to sanitize the water in the swimming pool that she and her husband owned. She had paid for the colchicine with a check for $21.20, which was later returned because it had been written on a closed bank account. With these new data the medical examiner’s office was notified, and a forensic pathologist reviewed the case.
A review of the postmortem findings used to substantiate the possible colchicine poisoning included hospital records, additional laboratory tests ordered, and a review of autopsy slides. The most conspicuous histologic changes were found in the tissues with rapid cellular turnover. The esophagus and stomach showed diffuse sloughing of the surface epithelium. The liver had early central lobular necrosis and areas of hepatocytes in arrested mitosis. Pseudo-Pelger anomaly (abnormal neutrophils) were present in the peripheral blood and could be seen in the blood vessels in the microscopic sections of tissue.
Fortunately, the serum from the original comprehensive drug screen was still available. The serum level for colchicine was 160 ng/ml, a lethal level. Wet tissue analysis of the kidney, liver, and esophagus was also performed for colchicine level determination. The following results were obtained: kidney, 160 ng/g; liver, 80 ng/g; and esophagus, 90 ng/g.
The wife was tried and convicted of intentionally poisoning her husband and was sentenced to prison.
The patient, a 35-year-old unmarried registered nurse, lived with another woman. One month before her death, she came to the local hospital with a history of diffuse abdominal pain, nausea, vomiting, and diarrhea; she attributed these symptoms to food poisoning. The patient recovered but had similar symptoms about 2 weeks later, which lasted 48 hours. She continued to experience midepigastric pain, for which she took two or three acetaminophen. On the evening before her last admission to the local hospital, her roommate found her unresponsive and intermittently vomiting bilious material. On admission, her temperature was 101°F and her blood glucose level was 32 mg/dl. She was given glucose and naloxone, with minimal results, and was then transferred to a nearby university medical center.
The medical history of this individual 1 year before hospital admission included pancreatitis and hepatic dysfunction, with negative results from a viral hepatitis panel. It also included endometriosis and frequent urinary tract infections. Her medications included omeprazole, promethazine, prochlorperazine, sucralfate, birth control pills, and pancrelipase. Her social history revealed that she was employed, had questionable alcohol use, and had smoked two packs of cigarettes per day for 10 years. There was no documented history of substance abuse.
On admission to the local university medical center, her vital signs were unremarkable except for a tachycardic pulse of 143/minute and shortness of breath. Examination of the abdomen revealed generalized guarding and right lateral upper quadrant pain. Neurologically, she could only grunt, could not fully cooperate, and would withdraw only to painful stimuli. Positive nuchal rigidity and positive deep tendon reflexes were also observed.
The admission laboratory data in hematology included a leukocyte count of 2.3 × 103/μl, hematocrit 25%, total bilirubin 3.0 mg/dl, serum glutamic oxaloacetic transaminase 142 IU/L, serum glutamic pyruvic transaminase 237 IU/L, lactate dehydrogenase 2025 U/L, alkaline phosphatase 91 IU/L, urea nitrogen 30 mg/dl, creatinine 2.0 mg/dl, lactic acid 5.5 meq/L, ammonia 295 μmol/L, and glucose 153 mg/dl. Toxicologic urine studies revealed no amphetamine, barbiturates, or benzodiazepines. However, dihydrocodeine and acetaminophen were identified. The results of serologic studies were positive for acetaminophen 20.8 μg/ml and negative for ethanol, methanol, isopropanol, and ethylene glycol. The results of a complete evaluation for infectious disease were negative except for a positive blood culture of Clostridium septicum, for which the patient was treated.
A gastroenterology consultation resulted in a tentative diagnosis of acute hepatic injury secondary to possible chronic alcohol abuse exacerbated by acetaminophen. The patient experienced adult respiratory distress syndrome, hepatic encephalopathy, and eventually disseminated intravascular coagulation. After the fourth day of hospitalization, her pupils became fixed and dilated. Emergency computed tomography revealed an intracranial hemorrhage. She expired 4 days after admission. Because of the unusual circumstances of her demise as well as suspicions on the part of family members that there might have been foul play, the medical examiner was contacted and accepted the case for autopsy.
At autopsy there was marked icterus. In addition, there were bilateral diffuse alveolar damage of lungs (right 1200 g, left 1150 g), a bright yellow firm liver with extensive hepatocellular necrosis (1700 g), moderate chronic pancreatitis with no gross lesions, right frontal lobe infarction of the brain with intracranial hemorrhage, and brainstem infarction with pontine hemorrhage. Microscopically, cells with pseudo-Pelger anomaly (abnormal neutrophils), including coarse clumping of chromatin, were seen in three separate organs: heart, spleen, and brain. Sections of esophagus and duodenum revealed vesiculated, hyperchromatic nuclei with clumped chromatin and mitotic figures in metaphasic arrest. In addition, in the esophagus there were cells that were no longer squamoid but more polygonal or globular. Dyskeratotic cells were also found in the esophagus. Both the proximal tubular cells of the kidney and the hepatocytes demonstrated similar nuclear findings as well as numerous arrested cells in metaphase. The bone marrow was hypercellular and, of all the organs examined, revealed the most cells in arrested metaphase. Extensive toxicologic testing revealed no drugs in the blood, and hydrocodone and dihydrocodeine in the urine. The result of a serologic test for cyanide was negative.
There was more than adequate pathologic evidence for a cause of death in this case. However, no infectious or toxicologic agent was identified that might have initiated the chain of events for the initial development of upper gastrointestinal symptoms and hepatic failure that eventually led to her death. The decedent had been scheduled to testify on behalf of a friend in a bitterly contested child custody case because she had been the only person present during an abusive situation between the warring parties. Extensive interviews and a search for a possible poisonous substance in and around the decedent’s apartment yielded no useful information.
Because of the family members’ concerns as well as the upper gastrointestinal symptoms and hepatotoxicity, the decedent’s university hospital admission blood specimen was analyzed for arsenic and colchicine. The specimen was sent to the National Medical Service, Inc., in Willow Grove, Pennsylvania. Arsenic was not detected, but the colchicine was 170 ng/ml, a lethal level. Further toxicologic studies of wet tissue of the liver, kidney, and brain revealed no residual colchicine. Investigation of the local drugstore revealed no evidence of any missing colchicine, nor did the State of Tennessee Pharmacy Board have records that this drug was missing from any source. The decedent did not have prescription colchicine and, according to her family, had never expressed suicidal ideations. After a lengthy investigation, the manner of death was ruled a homicide by colchicine poisoning. However, the person or persons who gave the decedent the drug have never been identified, and the case remains unsolved.
The potential hazards of industrial compounds, pesticides, dangerous household products, and other toxic substances in the environment have received wide recognition in the popular press. Toxins can be divided into nonpharmaceutical poisons and pharmaceutical substances. A poison is an agent that can seriously compromise function and/or cause death. Many prescribed pharmaceutical substances such as colchicine have recognized therapeutic ranges. However, the drug becomes a pharmaceutical poison when the therapeutic ranges are exceeded (7–13).
It is important to correlate clinical signs and symptoms with autopsy findings. Differential diagnoses in these two cases might have included drug or toxic substance ingestion, including arsenic, acetaminophen, lead, mushrooms, or other unknown substances. If the toxicology findings do not correlate with the clinical signs and symptoms, and the autopsy findings are nonspecific, then one must search for the unsuspected. A pharmaceutical poisoning should be considered when signs and symptoms suggest a drug or chemical and the results of toxicology screens are negative.
The circumstances and investigative information surrounding a death are crucial to cases of this type. In both these instances, autopsy findings were nonspecific, and the standard toxicology screens did not identify the toxic agent. However, when all the data from family members, other involved persons, and the illness trajectory were carefully correlated with the autopsy results, rare and unexpected findings resulted.
The authors thank Ms. Sherry Lewis and Ms. Diane Eigsti-Gerber for their assistance in editing and reviewing this article.
1. Kintz P, Jamey C, Tracqui A, et al. Colchicine poisoning
: report of a fatal case and presentation of an HPLC procedure for body fluid and tissue Analyses. J Anal Toxicol 1997; 21:70–2.
2. Folpini A, Furfori P. Colchicine
toxicity: clinical features and treatment: massive overdose case report. Clin Toxicol 1995; 33:71–7.
3. Neuss MN, McCallum RM, Brenckman WD, et al. Long-term colchicine
administration leading to colchicine
toxicity and death. Arthritis Rheumatism 1986; 29:448–9.
4. Gilman AG, Goodman LS, Rall TW, et al., eds. Goodman and Gilman’s the pharmacological basis of therapeutics,
8th ed. New York: Macmillan, 1990:674–6.
5. Bowder D. Who’s who in the Greek world.
New York: Washington Square Press, 1982:204–5.
6. Spillberg I, Mandell B, Mehta J, et al. Mechanism of colchicine
action in acute urate crystal-induced arthritis. J Clin Invest 1979; 64:775–80.
7. Mendis S. Colchicine
cardiotoxicity following ingestion of Gloriosa superba
tubers. Postgrad Med J 1989; 65:752–5.
8. Putterman C, Ben-Chetrit E, Caraco Y, et al. Colchicine
intoxication: clinical pharmacology, risk factors, features and management. Semin Arthritis Rheumatism 1991; 21: 143–55.
9. Murray SS, Kramlinger KG, McMichan JC, et al. Acute toxicity after excessive ingestion of colchicine
. Mayo Clin Proc 1983; 58:528–31.
10. Clevenger CV, August TF, Shaw LM. Colchicine poisoning
: report of a fatal case with body fluid analysis by GC/MS and histopathologic examination of postmortem tissues. J Anal Toxicol 1991; 15:151–4.
11. Klintschar M, Beham-Schmidt C, Radner H, et al. Colchicine poisoning
by accidental ingestion of meadow saffron (Colchicum autumnale
): pathological and medicolegal aspects. Forensic Sci Int 1999; 106:191–200.
12. Milne ST, Meek, PD. Fatal colchicine
overdose: report of a case and review of the literature. Am J Emerg Med 1998; 16: 603–8.
13. Mullins ME, Carrico EA, Horowitz BZ. Fatal cardiovascular collapse following acute colchicine
ingestion. J Toxicol Clin Toxicol 2000; 38:51–4.