Faculty Bibliography

Background: Amygdalin, marketed misleadingly by the misnomer "Vitamin B17", is a cyanogenic glycoside found in certain seeds (apricot, bitter almond, etc.). When swallowed, it is hydrolyzed in the small intestine into cyanide, and absorbed systemically. Despite a ban from the U.S. FDA and a Cochrane Review concluding no benefit to treat cancer, amygdalin is still available for purchase over the internet in 500mg capsules, each containing up to 30mg of cyanide. Thus, an estimated adult LD50 is as few as four capsules. We present a massive intentional overdose of amygdalin, with delayed recurrent hyperlactatemia, and successful combination antidotal therapy. Case report: A 33-year-old woman presented to the ED approximately 5 h after intentionally ingesting 40 capsules of 500mg amygdalin (20 g). On arrival, her vital signs were: HR 127/min, BP 112/65mmHg, RR 25/min, SpO2 98%. She was in agitated delirium, diaphoretic and mydriatic with an ECG notable for QTc 538ms. Lorazepam 2mg and magnesium sulfate 2 g were administered for agitation and prolonged QTc, respectively. Within 30 min of sedation, she became comatose, more tachypneic/diaphoretic, and hypotensive; therefore, 5 g hydroxocobalamin was empirically administered. Her pre-treatment venous blood gas showed: pH 7.27, pCO2 17.1mmHg, HCO3 7.7mmol/l, and lactate 14.1mmol/l. After initial treatment, she was noted to have oral foaming and hypoxia, so was intubated with sodium bicarbonate pre-treatment and another 5 g of hydroxocobalamin was given along with 25 g of sodium thiosulfate. Her BP normalized, QTc shortened, and lactate fell to 0.9mmol/l. Approximately 12 h later, her BP decreased to 60 s/40 s, her lactate increased to 8.1mmol/l and her QTc prolonged to 547 ms. Norepinephrine and vasopressin were started, activated charcoal was given and magnesium sulfate, sodium bicarbonate, hydroxocobalamin and sodium thiosulfate were redosed as previously with an improvement in vital signs, decrease in QTc and clearance of lactate. Her course was complicated by pneumonia, requiring continued sedation and ventilation, but she was extubated with no neurologic deficits on hospital day 9. Initial serum cyanide concentration was found to be 400mcg/l. Case discussion: Given that each 500mg capsule of amygdalin contains up to 30mg of cyanide, this patient ingested as much as 1.2 g of cyanide. The first-line antidote, hydroxocobalamin, has a binding ratio of 50:1 by molecular weight. So, for every 5 g of hydroxocobalamin, only approximately 100mg (4-5 capsules) of hydrogen cyanide will be sequestered. As might be expected, toxicity recurred due to ongoing absorption and/or decreased motility to the distal gut, where the majority of cyanide absorption may occur. Finally, we demonstrate the need to redose hydroxocobalamin and sodium thiosulfate in the case of recurrent cyanide toxicity from massive amygdalin overdose. Conclusions: Amygdalin remains a significant threat to human life with no role as a clinical therapeutic. Internet marketers should be held accountable by government authorities. We present a case of severe, recurrent cyanide toxicity secondary to massive amygdalin overdose, requiring multiple staggered doses of hydroxocobalamin and sodium thiosulfate

Background: Extracorporeal membrane oxygenation (ECMO) is a rescue therapy for patients with acute respiratory distress syndrome (ARDS) and refractory cardiogenic shock. Patients with diphenhydramine overdose can develop severe cardiotoxicity, including wide-complex tachycardia leading to cardiac arrest. There are no reported cases of confirmed diphenhydramine poisoning successfully treated with ECMO Hypothesis: ECMO is effective in severe diphenhydramine poisoning with ARDS and refractory cardiogenic shock. Methods: Single-patient chart review. Case: An 18-year-old female with a history of depression was brought to the emergency department (ED) after being found unresponsive in her car with an empty package of diphenhydramine and empty bottle of ibuprofen Initial emergency medical services (EMS) vital signs were BP 60 mmHg/palp, HR 114 bpm, and RR 12 bpm. Bag-valve mask ventilation was initiated en route to the hospital. In the ED, the patient was unresponsive and then began actively seizing. Cardiac monitoring revealed a wide-complex tachycardia, and then the patient went into a pulseless electrical activity (PEA) cardiac arrest Return of spontaneous circulation occurred after 6 min of CPR during which she was intubated and received intravenous sodium bicarbonate, epi-nephrine, dextrose, calcium, and normal saline. Despite vasopressors and maximum ventilator support, the patient developed ARDS and refractory cardiogenic shock. She was placed on veno-arterial (VA) ECMO. The patient received VA ECMO for 3 days, then veno-venous (VV) ECMO for 11 days, and mechanical ventilation foratotalof21 days. Her course was complicated by rhabdomyolysis, acute kidney injury requiring dialysis, acute liver failure, and compartment syndrome of her left lower extremity necessitating fasciotomy. She was discharged to inpatient rehabilitation neurologically intact 30 days after presentation. A serum diphenhydramine concentration obtained in the ED on arrival was 6000 ng/mL (50-100 ng/mL). Acetaminophen, salicylate, and urine toxicology testing were all negative. Discussion: We believe this is the first case of confirmed diphenhydra-mine poisoning successfully treated with ECMO. This case report supports the use of ECMO in poisonings with cardiovascular collapse secondary to a cardiac toxin. Conclusion: ECMO can be used as a life-saving treatment modality in severe diphenhydramine overdose refractory to conventional therapy

Background: In 2016, the American Association of Poison Control Centers/National Poison Data System (AAPCC/NPDS) recorded over one million exposure calls for children under 6-years-old with 12.8% requiring management in a healthcare facility. While widely used as a method of poisoning surveillance and epidemiologic information, the AAPCC/NPDS requires active reporting and is subject to significant underreporting of cases. Objective: The aim of this study was to assess the reporting of pediatric poisoning cases treated at healthcare facilities (HCF) to our regional poison control center (PCC) based on poison exposure category. Methods: This analysis builds on a previously unpublished retrospective review assessing factors associated with PCC reporting. We obtained poisoning visit information from two databases: PCC Toxicall and the Statewide Planning and Research Cooperative System (SPARCS) over a 4-year period (2010-2014). SPARCS is an all payer database that collects patient demographics, diagnoses, treatments, and charges for all inpatient hospital services and emergency department visits. Children under 6-years-old treated at a HCF located within the PCC primary catchment area with ICD9 diagnosis codes 960-979, 981-987, or 989 were identified in SPARCS. These cases were then uniquely matched to PCC Toxicall on patient age, sex, healthcare facility, name, and admission date. The resulting matched dataset was used to describe PCC reporting for visits within each poison category. ICD9 diagnosis codes were grouped into poison categories consistent with AAPCC/NPDS major categories. We calculated the proportions of each poison category reported to the PCC and determined which poison categories had highest and lowest percentages of reporting. We used descriptive statistics to characterize the data. Results: We identified 11,620 children in SPARCS under the age of six years treated in a local HCF with a poisoning diagnosis code; 59.8% (N=6951) of these patients were matched to a reported PCC Toxicall case. The poison categories with the largest percentage of cases reported to PCC included: anticonvulsants (N=73, 84.9% reported; RR 1.45, p<001), cardiovascular drugs (N=420, 80.7% reported; RR 1.40, p<001), antihistamines (N=336, 79.8% reported; RR 1.37, p<001)), antidepressants (N=115, 79.1% reported; RR 1.35, p<001), and sedatives/hypnotics (N=405, 77.8% reported; RR 1.34, p<001). Of the unreported cases (N=4669), the poison categories with the smallest percentage of cases reported to PCC included: venom (N=755, 0.3% reported; RR 0.004, p<001), heavy metals/lead (N=117, 1.7% reported; RR 0.03, p<001), CO/gases (N=972, 15.4% reported; RR 0.25, p<001), plants/mushrooms (N=84, 16.7% reported; RR 0.28, p<001), and antimicrobials (N=183, 56.3% reported; RR 0.97, p=.62). Conclusions: Environmental exposures, including heavy metals/ lead and carbon monoxide, were associated with a high percentage of underreporting to our regional PCC. These findings suggest an area for focused outreach and education regarding the availability of the PCC in the consultation and treatment of all types of pediatric poisoning exposures, specifically including non-medicinal/environmental exposures

Background: In 2016, the American Association of Poison Control Centers (AAPCC) recorded over one million exposure calls for children under 6-years-old, with 12.8% requiring management in a healthcare facility. Despite poisoning being a mandatory reportable condition in our jurisdiction, underreporting to our PCC persists. Objectives: The aim of this study was to determine the factors associated with healthcare facility (HCF) reporting of childhood poisoning cases to our PCC. Methods: This study is a retrospective database review using 2010-2014 poisoning visit information from two databases: PCC Toxicall and the Statewide Planning and Research Cooperative System (SPARCS). SPARCS is an all payer database that collects patient demographics, diagnoses, treatments, and charges for all inpatient hospital services and emergency department (ED) visits. Children under 6-years-old presenting to a HCF in the PCC primary catchment region with ICD9 diagnosis codes 960-979, 981-987, or 989 were identified in SPARCS. These cases were then uniquely matched to the PCC Toxicall data based on patient age, sex, HCF, name, and admission date. Matched cases were then classified as poisoning visits reported to PCC. Chi-square statistics were used to assess the univariate association between PCC reporting and age, sex, poverty level, county, private insurance, hospital type (with or without a pediatric intensive care unit-PICU), presenting location (ED versus inpatient), disposition (admit to ICU, admit to floor, or discharge), length of stay, or poison category (drug/pharmaceutical versus non-medicinal/environmental). We then used stepwise logistic regression to model the relationship between these covariates and PCC reporting. Relative risks with corresponding 95% confidence intervals were then approximated from the adjusted odds ratios. Results: We identified 11,620 children in SPARCS under the age of six years treated in a local HCF with at least one poisoning diagnosis code. Only 59.8% of these patients were matched to a unique PCC Toxicall case, indicating underreporting by 40.2%. Young age, having private insurance, presenting to the ED, being treated at a hospital with a PICU, admission to ICU, length of stay longer than 24 h, being treated for a drug/pharmaceutical poison and treatment within the same county where the PCC is located were all associated with increased reporting to the PCC. After controlling for age, county, insurance type, poison type, disposition, location of presentation, and length of stay, younger children (age <1 years; RR=1.35, 95%CI 1.28-1.41) and those with drug/medicinal exposures (RR 1.62, 95%CI 1.58-1.65) were more likely to be reported to PCC. Additionally, patients transferred for care with a resultant length of stay longer than 24 h were more likely to be reported to PCC (RR=2.86; 95%CI 1.6-4.01). Conclusions: Only 60% of pediatric poison exposures were reported to our PCC by healthcare facilities. While, the youngest patients and those that require transfer for longer care are more likely to involve the PCC for consultation, additional outreach and education is required to improve PCC penetrance