Iron Poisoning

Jeffrey S. Fine, MD

D ron poisoning has always been of particular Iron is a group-VIII transition metal, with electrons interest to pediatricians: children are frequently distributed throughout five 3d orbitals. When iron forms exposed to iron-containing products, and they chemical compounds or complexes with large proteins experience the worst toxicity. The first well-described such as hemoglobin, these electrons shift between cases of iron poisoning were published in the 1940s higher and lower energy orbitals and achieve different and 1950s, the classic animal experiments were per- spin states. 2,4 With this variable electron distribution, formed in the 1950s and 1960s, and the use of the iron can exist in oxidation states between -2 and +6, chelator deferoxamine as an antidote for iron poison- either donate or receive electrons, and achieve variable ing was introduced in the 1960s. redox potentials during electrochemical reactions. Most of what we know about the clinical presentation, This makes iron an ideal mediator of diverse biologi- the pathology, and the pathophysiology of iron poison- cal redox reactions such as when electrons are trans- ing comes from this early work although research con- ferred down the mitochondrial cytochrome chain. tinues. What has changed most over the past 50 years is Table 1 lists some basic facts about the physical chem- the approach to management. As with many types of istry and biochemistry of iron. poisoning, there is now an emphasis on excellent sup- Physiologically, iron exists primarily in 2 stable oxi- portive care with an individualized approach to gastroin- dation states: ferrous iron (Fe +2 or Fe [II]) and ferric testinal decontamination and a selective use of antidotes. iron (Fe +3 or Fe [III]). In aqueous solution Fe +2 is oxi- With this strategy, the deaths related to iron poisoning dized to Fe +3, which often forms insoluble complexes have been reduced to only a small number each year. This at physiologic pH. Iron chelators such as transferrin or article reviews the epidemiologic, clinical, animal, and deferoxamine form soluble complexes with Fe +3. laboratory science related to iron poisoning and its man- A typical American diet contains approximately 15 agement and focuses on several areas of controversy. to 40 mg of iron per day most of which is in the ferric form, which is insoluble at physiologic pH and poorly absorbed. This Fe +3 is released from food proteins and Background reduced to the ferrous form by gastric acid. Iron is the fourth most abundant atomic element in Absorption of dietary iron takes place primarily in the the earth's crust and is found in the minerals hematite, duodenum and upper jejunum and is regulated accord- magnetite, and siderite.~ Biologically, iron is an essen- ing to the state of the body's iron stores. Heme iron is tial nutrient for most living organisms because it is a absorbed better than nonheme iron, and different food component or cofactor of many critical proteins and components may affect iron absorption. For instance, enzymes such as hemoglobin, myoglobin, catalase, ascorbate increases and phytates decrease absorption; xanthine oxidase, aconitase, reduced nicotinamide however the overall effect of food is to decrease iron adenine dinucleotide, ribonucleotide reductase, perox- absorption. The body efficiently recycles iron released idases, and cytochrome oxidase. 2,3 from senescent red blood cells, so the daily requirement for iron is approximately 1 mg to cover losses of iron in stool, sweat, and urine. Menses accounts for an

Jeffrey S. Fine, MD, is Assistant Director of Pediatric Emergency additional need of 1 mg/d for women. Thus approxi- Services at Bellevue Hospital and an assistant professor of Pediatrics mately 10% of dietary iron is absorbed. 1-3 at New York University School of Medicine, New York, New York. The intestinal mucosal cell has few barriers to iron Curr Probl Pediatr 2000;30:71-90. Copyright © 2000 by Mosby, inc. uptake, but once inside the cell the transfer of iron to 0045-9380/2000/$12.00 + .15 53/1,/104055 the plasma is highly regulated. The physiology of iron doi:10.1067/mps.2000.104055 absorption and metabolism is quite complex and out-

Curr Probl Pediatr, March 2000 71 TABLE 1. Iron facts Epidemiology of Iron Poisoning

Atomic no.: 26 The American Association of Poison Control Centers Atomic weight: 55.85 (AAPCC) is a consortium of 66 regional poison control Density: 7.784 g/cm 3 and information centers located throughout the United Normal serum level: 80-180 pg/dL (1¢32 #mol/L) States. Each year the AAPCC collects standardized epi- Ferrous = Fe+2 or Fe (11) demiologic data about poisonings reported to these cen- Ferric = Fe+3 or Fe (111) ters, and this information is published annually. For Required daily dose: I mg/kg per day (maximum 15 mg) 1997, the AAPCC reported 1.5 million poisoning exposures for children and adolescents younger than 20 years. 9 These pediatric exposures accounted for 67% of side the scope of our discussion; the reader is referred all the reported poisoning exposures. Children younger to several recent reviews for more information. 1&5 A than 6 years accounted for 77% of the pediatric expo- typical adult has approximately 3 to 4 g of iron (30-40 sures and 50% of all reported exposures. mg/kg body weight); most of this iron is found in In 1997, there were 26,544 AAPCC-reported expo- hemoglobin or stored as ferritin. Only 3.5 mg (0.1%) sures for all types of iron-containing compounds by circulates in the plasma. 2 children and adolescents accounting for 1.7% of all Iron toxicity is related to the generation of free rad- reported pediatric poisoning exposures; 87% were in icals. 6,7 Both normal and pathologic cellular process- children younger than 6 years. Sixty-two percent of es produce superoxide (02-) and hydrogen peroxide these childhood exposures were to pediatric multivita- (H202) byproducts, whereas enzymes such as super- min tablets, which rarely lead to serious toxicity after oxide dismutase, glutathione peroxidase, and catalase ingestion. Thirty percent of these exposures were to normally metabolize and neutralize these free radicals. adult preparations such as iron pills or prenatal vita- One of the effects of 02 - is the release of stored iron mins with iron, which are responsible for almost all of from ferritin. The free iron reacts with O 2- and H202 the serious toxicity related to iron poisoning. to produce other more reactive and toxic-free radicals Iron-containing products are available in many house- such as the hydroxyl radical: 6,7 holds with young children. Iron pills are prescribed for most pregnant women, and many children are receiving 02- + Fe +3 --~ 02 + Fe +2 multivitamins with iron. Many of the prenatal vitamins are brightly colored, sugar coated, and resemble popu- Fe +2 + H202 --~ Fe +3 + HOo + OH- lar candies, such as M&M's or Good & Plenty (Fig 1). 1° Many children's multivitamins are associated with car- O2o- + H202 --~ O 2 + HOo + OH- toon characters like the Flintstones. In this regard, the The hydroxyl radical (HOo) formed through this comments of Spencer, u written in 1954, seem timely: reaction can depolymerize polysaccharides, cause Now it seems irrational that some things which children are DNA strand breaks, inactivate enzymes, and initiate expected to take, such as cod liver oil, remain unpalatable, lipid peroxidation, which is a self-amplifying process whereas ferrous sulphate tablets, which are intended main- that is particularly damaging to cellular and subcellular ly for adult use, are made both attractive and sweet. membranes. 6,7 For example, when lysosomal mem- The reasons that children ingest drugs are multifactor- branes are destroyed, proteolytic enzymes are released ial and include developmental level, inadequate parental inside the cell, causing further cell damage. If the dam- knowledge about the toxic potential of individual agents age is irreparable, it can also lead to cell death. 8 such as iron, access to medications, imitative behavior, The body protects itself from the toxicity of free iron lack of supervision, and intrafamilial conflict.12,13 by keeping it chelated during almost all phases of Almost all deaths from iron poisoning are in children absorption and distribution. From the enterocyte, iron is younger than 3 years and follow large exposures to adult accepted for transport through the plasma by transferrin, iron preparations (30-40 pills), although death after the a [31-glycoprotein with 2 high-affinity binding sites for ingestion of as few as 5 pills has been reported. 14,15 Fe+3. One third of the body's iron is stored as ferritin, a Early series reported mortality rates of 50% from iron macromolecule capable of binding up to 4500 iron poisoning, which undoubtedly reflected a reporting bias atoms. Sixty percent of this ferritin is in hepatocytes, and for very sick, hospitalized patients. 16 Despite the extra- 40% is in myocytes and reticuloendothelial cells. ordinary number of exposures to iron-containing prod-

72 Curr Probl Pediatr, March 2000 ucts, deaths rarely occur. For the past 10 years, there has been an average of only 3 to 4 deaths per year reported by the AAPCC, although there are sporadic increases; in 1992, 11 childhood deaths were reported.14,17 This number of reported deaths may be an underestimation of the actual incidence, because some deaths are not reported to regional poison control centers. 18 Although mortality statistics are relatively well maintained, data about poisoning morbidity are difficult to ascertain. The AAPCC reports a significant amount of epidemiologic information related to poisoning, but the annual reports do not stratify all the data according to age. However, the AAPCC did a spe- cial review of morbidity and mortality in children younger than 6 years for the period 1985-1990.19 Eight of 111 deaths during this period were related to FIG 1. Iron sulfate tablets alongside popular candies. Can you tell the iron poisoning; an additional 85 children had "major" difference? (Reprinted with permission from Goldfrank LR, Flomenbaum NE, Lewin NA, Weisman RS, Howland MA, Hoffman life-threatening morbidity. Again, this may be an under- RS, editors. Goldfrank's toxicologic emergencies. 5th ed. Stamford: estimation of the number of children who experience Appleton & Lange; 1994. p. 522.) severe iron poisoning. Several reviews of poisoning hospitalizations suggest that iron poisoning is an infre- TABLE 2. Iron-containing preparations quent cause of hospitalization2°-23; the hospitalization Tablets or capsules rate for iron poisoning among children aged 0 to 4 mg Fe+2 per tablet years has been estimated at 8.7 per 100,000 children.24 Preparation Example or capsule Carbonyl iron Feosol Caplet 50* Ferrous fumarate Femiron 20 Prevention Feostat 33 Chromagen 66 A representative sample of iron-containing prepara- Hemocyte 106 tions is listed in Table 2. One single 324-mg ferrous sul- Ferrous gluconate Fergon 27 fate tablet will cause gastrointestinal symptoms in 10% Generic 36 of adults who take iron supplements, and 3 to 4 pills rep- Ferrous sulfate Slow-Fe 50 Feosol tablet 65 resent a potentially toxic dose in a toddler (Table 3). One Fero-folic-500~ 105 way to reduce the frequency of significant poisoning would be to decrease the concentration of iron in each Liquid preparations pill. Children's multivitamins contain only 15 to 18 mg Preparation Example mg Fe+2 of elemental iron per tablet, and this accounts for the Ferrous fumarate Feostat suspension 33/5 ml lower frequency of severe poisonings related to these Feostat drops 15/0.6 ml Ferrous sulfate Fer-ln-Sol drops 15/0.6 ml products. However, reducing the quantity of iron in adult Fer-lr~Solsyrup 18/5 ml tablets would require increasing the number of daily doses. Feosol elixir 44/5 ml Although many modalities may help to diminish the *Carbonyliron is Fe°. incidence of childhood poisonings in general and iron l-Fe+2 in multivitaminsranges from lOto 200 mg per tablet and 4to 100 mg/5 ml. poisoning in particular, the surest method is to limit access by keeping children physically separated from mg of iron per container be in child-resistant packag- iron-containing products. Again, almost 50 years ago ing. Unfortunately, child-resistant is not childproof; Spencer 11 recommended that containers may be left open, the closures may not the daily dose of three tablets should be wrapped separate- function correctly, or a sibling may open the contain- ly in small packets in the same way that sugar is sometimes er. In 1997 the Food and Drug Administration (FDA) supplied in hotels in packets containing two lumps. issued a regulation requiring products that contain 30 In 1987, the Consumer Product Safety Commission mg or more of iron per dosage unit to be packaged as required that food supplements with more than 250 individual doses (for example, blister packs). 25

Curr Probl Pediatr, March 2000 73 TABLE 3, Guidelines for predicting toxicity and the need for treatment in a 12-kg child No. of tablets for Evaluation* Treatment* Preparation % Fe+2/tablet mg Fe+2/tablet (20 mg Fe+2/kg) (60 mg Fe+2/kg) Ferrous sulfate 20 65 4 12 110 2 6 Ferrous fumarate 33 66 4 12 110 2 6 *These are poison center triage guidelines. See Table 6,

TABLE 4. Stages of acute iron toxicity nat effects are attributed to the direct local corrosive Stage I: local gastrointestinal effects effects of iron on the gastric and intestinal mucosa and Stage I1: quiescent phase occur early, usually within several hours. In animal Stage II1: systemic toxicity Stage IV: recovery models and in human patients who have required Stage V: gastrointestinal obstruction surgery or died, pathologic changes range from mild congestion to hemorrhagic necrosis of the intestinal mucosa, and an iron-filled exudate or coating of the Carbonyl iron is another form of iron that has been mucosa is often seen. 11,31-34 Although pathologic reintroduced over the past 15 years for the treatment effects are concentrated in the stomach and duode- of iron-deficiency anemia, mostly in Europe. Carbonyl num, areas of small bowel infarction can be observed iron is produced when iron pentacarbonyl is vaporized distally; the colon is rarely affectedY ,36 These initial and crystallized into submicroscopic spheres of local corrosive changes sometimes lead to the forma- uncharged highly purified metallic iron (Fe°). This tion of gastric antral and pyloric strictures, which are form of iron must be converted to Fe 2 by gastric acid observed in some patients between 2 and 8 weeks after before it can be absorbed, and as a result, carbonyl iron the initial episode of poisoning (Stage V). 36-38 is absorbed more slowly than ferrous sulfate, although Stage II is a quiescent phase during which there is the bioavailability of the absorbed iron is high. 26 resolution of gastrointestinal symptoms with apparent The typical adult dose of carbonyl iron is 100 mg/d, clinical improvement. Some of the first reported cases and volunteers have been given up to 10,000 mg/d of iron poisoning described children who did not (140 mg Fe°/kg). 27,28 The gastrointestinal adverse receive any specific therapy, had resolution of gas- effects of even the high dose are similar to those of trointestinal symptoms, and became critically ill after ferrous sulfate, although there is more diarrhea and the 24 hours and died. 31,33 The existence of this phase is carbonyl iron has a funny metallic taste. The highest controversial because many toxicologists believe that dose at which no deaths occur in animals (LDo) is esti- patients who are significantly poisoned will almost mated at 10,000 to 15,000 mg Fe°/kg, whereas the always become sick early. From the limited clinical median lethal dose (LDso) for ferrous sulfate is esti- data presented in those early reports, it is difficult to mated at 250 mg Fe2/kg. 29 These data suggest that assess what the true clinical status of the children limited absorption prevents carbonyl iron toxicity and was before and during the quiescent phase or whether that unintentional ingestion will not result in serious death may have been related to another cause, such as injury or death. Therefore, the FDA has temporarily aspiration. The possibility of the quiescent phase is exempted carbonyl iron from the unit-dose packaging important to consider when a child is being managed requirements described above. 15 with observation alone; however, this stage would rarely be seen today because children with significant early gastrointestinal symptoms would receive fluid Pathophysiology of Iron Poisoning resuscitation and chelation therapy soon after presen- The clinical progression of iron poisoning is usually tation. divided into 5 stages (Table 4). Stage I is gastroin- Stage III represents systemic toxicity, which defines testinal toxicity; nausea, vomiting, and diarrhea have true iron poisoning, although a specific dose for this been observed in volunteers at doses as low as 5 mg effect has not been defined. Systemic toxicity is clini- Fe+2/kg.3° Gastrointestinal bleeding may lead to cally manifest as shock with associated signs of hematemesis or bloody diarrhea. These gastrointesti- hypoperfusion--pallor, cold extremities, tachycardia,

74 Curr Probl Pediatr, March 2000 TABLE 5. Serum iron levels after oral doses of iron Dose Peak serum iron level Time to peak Study* Formulation (mg Fe+2/kg) (gg/dL) (h) A3° Ferrous sulfate chewable multivitamins 5 243 4.2 B63 Ferrous fumarate chewable multivitamins 6 263 4 Ferrous fumarate tablets 6 125 6 A3° Ferrous sulfate chewable multivitamins 10 321 4.5 C64 Ferrous sulfate tablets 20 300 2-3 *Studies in adult volunteers and tachypnea; hypotension may be a late find- ritin. Pathologic changes range from cloudy swelling to ing. 32,34,39-41 Hypovolemic shock and acidosis are periportal hepatic necrosis, and elevated transaminases the primary determinants of this shock state, but are observed. 5°'51 Iron is concentrated within hepatic diminished cardiac function contributes (see below). mitochondria where it destroys mitochondrial mem- Hypovolemic shock results from the gastrointestinal branes through the free radical mechanisms described fluid losses described above as well as increased cap- above and disrupts the electron transport of oxidative illary permeability and possible loss of venous tone. 42 phosphorylation.52-54 These effects contribute to both Myocardial dysfunction may be seen in severe iron hepatic necrosis and metabolic acidosis. poisoning. In animal models, decreased heart rate, Abnormalities in coagulation have been observed in decreased mean arterial blood pressure, decreased car- animal models and some human patients with iron diac output, and increased peripheral resistance have poisoning; 32,39,55,56-58 increased bleeding may exacer- been observed even when intravascular volume and pH bate hypovolemia. Initially, iron may exert a direct were corrected. 4°,41,43,44 .Decreased cardiac output is inhibitory effect on clotting factors such as throm- related to decreased myocardial contractifity, which is bin. 57 Later, reduced levels of clotting factors may be exacerbated by acidosis and hypovolemia. 45,46 an effect of hepatic failure. 5s Although heart failure is frequently seen in chronic Stage IV is the period of clinical recovery that will iron overload, it has only been rarely reported after begin soon after the initiation of fluid and antidotal acute iron poisoning and may be related to free radi- therapy. For severely poisoned patients recovery will cal-induced damages In vitro, increased lysosomal be marked by resolution of acidosis and other signs of enzyme activity has been associated with iron loading shock, usually within 3 to 4 days of the acute poison- of cultured myocytes, and the sarcolemmal membrane ing, although full recovery may take longer. seems to be particularly susceptible to free radical lipid Stage V is the late onset of gastric and pyloric peroxidation. 45,46 In addition, inhibition of cardiac strictures, which may occur 2 to 8 weeks after the mitochondrial respiratory enzyme activity caused by initial injury. These strictures occur in mucosa in iron loading contributes to myocardial dysfunction.48 which there was previous damage, and they frequent- Systemic iron poisoning is associated with a posi- ly require surgical therapy. 36-38 tive anion-gap metabolic acidosis. 33,34,39-41 Several possible explanations for the acidosis have been proposed: (1) conversion of free plasma iron to ferric Iron Absorption in Overdose hydroxide is accompanied by a rise in hydrogen ion As described above, iron absorption is a highly regu- concentration; 39 (2) free radical damage to mito- lated process. How these processes are affected in over- chondrial membranes prevents normal cellular respi- dose is unknown. 59 It is clear that there can be rapid ration and electron transport with the subsequent absorption through intact intestinal mucosa. 32,34,60 development of lactic acidosis; 49 (3) hypovolemia There have been several volunteer studies that exam- and attendant hypoperfusion contribute to (but are ined iron absorption after supratherapeutic oral doses not the primary causes of) acidosis; 44 and (4) cardio- (Table 5). 30'61'62 They demonstrate that supratherapeutic genic shock contributes to hypoperfusion. 44,47 doses of iron are well absorbed after ingestion. A peak In some cases hepatic injury will be evident after 2 to serum iron level is achieved within several hours of 3 days. Portal blood delivers high concentrations of iron ingestion, and a level measured at 4 hours will be at or to the liver where it is taken up by both Kupffer's cells near the peak, although the peak level may be delayed. and hepatocytes and exceeds the storage capacity of fer- It is unclear whether very high doses lead to delayed

Curr Probl Pediatr, March 2000 75 I-I~N

ItO0

FIG 2. Straight-chain structure of deferoxamine. (Adapted with permission from Goldfrank LR, Flomenbaum NE, Lewin NA, Weisman RS, Howland MA, Hoffman RS, editors. Goldfrank's toxicologic emergencies. 6th ed. Stamford: Appleton & Lange; 1998. p. 628.)

be relatively asymptomatic. It may be that these iron bezoars reduce the rate of absorption because the iron inside the mass is not readily accessible for J (CH2)sNH2J absorption. If left untreated, these patients may experience severe delayed toxicity.

Deferoxamine (Desferrioxamine, ¢;°%/ Desferal) 63-65 Ferrioxamine B is an iron-containing siderophore ( 2)5-ff \'b "y" \(, :a2)= required by the fungus Streptomyces pilosus for normal \c" '=o growth. When the iron is removed, deferoxamine is produced. The straight-chain deferoxamine molecule folds around a single atom of iron to form a very stable octahedral complex, which occupies all of iron's reactive coordination sites (Figs 2 and 3). Deferoxamine has a FIG 3. Octahedral structure of ferrioxamine. (Reprinted with permis- very high binding constant for iron (1031) and, in vitro, sion from Goldfrank LR, Flomenbaum NE, Lewin NA, Weisman RS, removes iron from transferrin and ferritin but not from Howland MA, Hoffman RS, editors. Goldfrank's toxicologic emer- hemoglobin. In vivo, deferoxamine cannot remove iron gencies. 6th ed. Stamford: Appleton & Lange; 1998. p. 628.) from transferrin and probably chelates iron in transit between transferrin and ferritin. One mole of deferoxamine (100 mg) binds 1 mole of iron (9 rag) to form fer- absorption and peak level; however in many reported rioxamine, a reddish colored compound. cases, children became toxic soon after ingestion. Deferoxamine is administered through intravenous Although the peak level may be delayed, high serum or intramuscular injection because it has poor oral iron levels and the onset of toxicity are seen early. In absorption. The volume of distribution of deferoxa- animals given lethal doses of iron, near-peak levels are mine (0.6-1.2 L/kg) is greater than that of ferrioxa- observed within an hour. 32,34,44 mine (0.2 L/kg), which suggests that deferoxamine Iron pills can become conglomerated into a mass may distribute to tissues while ferrioxamine is con- in the stomach. In many of these cases, despite the fined to the vascular space. In fact, deferoxamine can observation of a large number of pills at gastrotomy remove iron from mitochondria and prevent iron- or on a radiograph, the patients are often reported to induced damage in that organelle. 66

76 Curr Probl PedJatr, March 2000 Deferoxamine was initially advocated for use as a gas- The most common complication seen with intra- tric lavage solution despite the fact that huge doses are venous administration of deferoxamine is hypoten- required to complex a significant oral iron dose: 100 mg sion. In the early animal models, deferoxamine was Fe+Z/kg in a 10-kg child would require 20 vials of defer- given at doses of 40 to 180 mg/kg per hour, or 3 to oxamine. The first animal models showed that guinea 10 times higher than the recommended dose, and pigs given an LDso dose of Fe+2 followed by oral defer- hypotension was common. 34 Increased levels of his- oxamine had increased survival compared with con- tamine measured during periods of hypotension may trols. 67 After this, the use of oral deferoxamine either contribute to the effect although pretreatment with alone or in combination with parenteral deferoxamine antihistamines does not ameliorate the effect. 77 was recommended as therapy for iron poisoning. Hypotension can occur at a dose of 15 mg/kg per Several groups reported its use without significant toxi- hour, and some clinicians recommend starting at a city although not all patients were severely poisoned and lower dose, for instance 8 mg/kg per hour, and titrat- most patients received both oral and intravenous defer- ing upwards until the desired dose is achieved. oxamine.68-71 Decreased iron absorption was observed Flushing reactions are also associated with deferox- in swine after receiving 60 mg/kg Fe +2 followed by oral amine administration and are considered an anaphy- deferoxamine72; however there was no decrease in iron lactoid reaction. absorption when human volunteers ingested 5 mg/kg The most concerning possible adverse effect of Fe+2 followed by oral deferoxamine.73 deferoxamine is the adult respiratory distress syn- The use of oral deferoxamine was questioned after drome (ARDS), which was observed in 4 cases of iron an experiment in which 4 dogs were administered poisoning in which deferoxamine was used. 78 These near lethal doses of 225 mg Fe+Z/kg as the iron- adult patients received 15 mg/kg per hour for 32 to 72 deferoxamine complex ferrioxamine. 34 The iron was hours, had symptoms of respiratory distress, and died. absorbed, and all of the dogs died. In the same study, Although this complication was associated with pro- 9 dogs were given both oral and intravenous defer- longed use of deferoxamine, 10 other patients with oxamine as an antidote for induced severe iron poi- deferoxamine infusions greater than 24 hours' dura- soning; only 3 of 9 dogs survived. Despite its appar- tion and 43 patients with infusions less than 24 hours' ently successful or at least nontoxic use in humans duration did not have symptoms. There is another case and some animal models, the use of oral deferoxa- report of a child with iron poisoning who had respira- mine is now rarely reported, and most clinicians have tory distress while taking deferoxamine.79 There are abandoned it as a lavage solution. 74 also several case reports of children with hemochro- The use of parenteral deferoxamine as an antidote for matosis or advanced malignancies in whom ARDS- iron overdose was derived from the use of deferoxa- like syndromes developed while they received defer- mine as therapy for transfusion-related iron overload. oxamine infusions for days or weeks, s°-82 Although a In this setting, intravenous deferoxamine led to relationship between ARDS and deferoxamine has not increased iron elimination compared with intramuscu- been definitely established, an animal model of defer- lar administration. 64 The efficacy of deferoxamine as oxamine-induced pulmonary toxicity has been created an antidote has never been tested in a human trial. 63 in hypoxic mice. Lung damage might be the result of However during those years that deferoxamine was inappropriate chelation of intracellular iron with inhi- becoming standard therapy, clinical outcomes improved bition of key enzymes or of iron-induced free radical compared with earlier case series. The current recom- damage as described above. 83 mended dose of 15 mg/kg per hour was determined Other adverse effects such as ocular damage and empirically and has also never been tested. 7t ototoxicity are associated with the chronic use of Limited data suggest that iron remains in the deferoxamine for transfusion-induced iron overload. 84 intravascular space for 12 to 24 hours before distrib- In addition, infections with Yersinia enterocolitica, ution is complete. It may be during this period that Zygomycetes, and Aeromonas hydrophilia and other iron is most available for chelation. Therefore, some organisms are also associated with the chronic use of toxicologists have proposed using higher doses of deferoxamine,85 although several Cases of acute iron intravenous deferoxamine during the initial treat- poisoning have been complicated by Y enterocolitica ment phase, and doses of 25 and 45 mg/kg per hour sepsis. 86-s8 Deferoxamine makes iron available to have been reported. 64,75,76 these organisms for enhanced growth.

Curr Probl Pediatr, March 2000 77 TABLE 6. Poison center referral guidelines where a small number of pills were reportedly ingest- Estimated dose (mg Fe+2/kg) Action ed, severe toxicity developed, suggesting that toxicity <20 Observe at home may not be related to dose. Individual factors related 20-60 Medical evaluation to absorption, iron-binding capacity, or individual sus- >60 Chelation therapy ceptibility must play some role in determining toxicity. However, because toddler ingestions are usually not witnessed, the most likely explanation for the dis- Deferiprone 89-90 crepancy between the reported iron dose and the severity of illness is inaccurate information about the Deferiprone (1,2-dimethyl-3-hydroxypyridin-4-one) ingested dose. is a relatively new, orally available iron chelator. Children who ing.est large numbers of pills often Whereas deferoxamine binds iron in a 1:1 molar ratio, become sick within 2 hours of ingestion or before 3 molecules of deferiprone are required to bind 1 atom parents seek medical attention. It is difficult to of iron. Thus relatively high doses of deferiprone are understand how a patient with a large number of required to effectively chelate and remove excess iron. pills visible on radiograph can have few symptoms, The primary adverse effects of deferiprone are transient but counting pills on radiograph has never been stan- agranulocytosis (0%-2%), transient muscutoskeletal dardized and may not accurately estimate the ingest- and joint pains (0%-30%), gastric intolerance (0%-6%), ed dose. Also, large numbers of pills may form an and zinc deficiency (0%-2%). Deferiprone is being iron bezoar that has a reduced rate of absorption evaluated for use in the treatment of iron overload states with delayed toxicity. for patients who cannot tolerate parenteral deferoxa- Because so many children are exposed to iron-con- mine or for patients who do not have access to par- taining products and so few have more than mild gas- enteral drug administration (eg, patients living in rural trointestinal irritation, there is a need for guidelines areas of developing countries). Its effectiveness for that can be used to decide which children with report- long-term therapy is under investigation. 91,92 ed ingestions require medical evaluation. The triage Deferiprone has also been considered as a possible guidelines shown in Table 6 were developed from the therapeutic agent for the treatment of acute iron poi- case literature and volunteer studies. soning and has been tested in animal models. Rats The absorption studies described above, in which poisoned with an LDs0 dose of Fe +2 had decreased the ingested dose is accurately known, show that tissue iron accumulation and lower mortality after adults with even small ingestions will have at least treatment with intraperitoneal deferiprone.93 However, mild symptoms. Volunteers who ingested 5 to 6 mg/kg mice poisoned with Fe +2 and treated with deferiprone of either ferrous sulfate or ferrous fumarate experi- experienced worse toxicity compared with controls. 94 enced some degree of nausea, diarrhea, or darkened At this point, deferiprone cannot be recommended for stools. 3°,61 All 6 volunteers who received 20 mg/kg the treatment of acute iron poisoning. ferrous sulfate had nausea, cramps, and diarrhea; however only one vomited. They all received intravenous fluid therapy and improved. 62 Although all these vol- Ancillary Information to Help Predict unteers experienced some gastrointestinal symptoms, Toxicity none had serious systemic clinical toxicity. Considering the high incidence of exposure to iron- Therefore, the action level of 20 mg/kg, which rep- containing products among young children, it would resents the ingestion of only a few pills by a small be helpful to have some useful tools or guidelines to child, should identify any patient with symptoms help management decisions. A number of different requiring medical evaluation (Table 4). It is important evaluation schemes and ancillary tests have been sug- to remember that these poison center-based triage gested over the years for this purpose. The following guidelines are based on the history of ingestion, sections review the utility of some of these tests. which is frequently inaccurate as described above. Therefore, any child with symptoms should also be Does the Estimated Dose Predict Toxicitye. referred for evaluation, regardless of the estimated Published cases usually report details about the pre- ingested iron dose. sumed number of iron pills ingested. In some cases Do these triage criteria work? In an early review of

78 Curr Probl Pediatr, March 2000 34 patients, 65% of patients believed to have ingested TABLE 7. Serum iron level for pediatric iron-poisoning deaths, less than 50 mg Fe+2/kg had mild symptoms or were 1987-1997 asymptomatic, whereas 88% of patients believed to Serum Fe ÷2 (gg/dL) No. have ingested more than 100 mg Fe÷2/kg were moder- < 300 2 300-500 3 ately or severely poisoned. 95 More recently, almost 500-1000 2 200 patients were stratified according to their estimat- 1000-5000 15 ed ingested dose. 96 Twenty-two percent of the group >5000 14 who took 20 to 40 mg/kg, 42% of the group who took Total 36 Data collated from annual reports of the AAPCC, 1988-1998, published in 40 to 60 mg/kg, and 33% of the group who took more the American Journal of Emergency Medicine each September. than 60 mg/kg had symptoms, although the highest reported serum iron level in any group was only 539 gg/dL (moderately elevated; see next section). No ity of poisoning. There are several possible reasons for patient had serious toxicity. According to these results, the discrepancy. the authors recommended home observation for The ideal serum iron level is a peak level drawn patients with an estimated dose of less than 60 mg/kg between 2 and 6 hours after ingestion (see "Iron Fe +2. Although this management strategy will general- Absorption" above). One of the problems with evalu- ly be safe, a low estimated dose does not guarantee a ating whether the serum iron level is a good predictor true low dose or a nontoxic ingestion. of toxicity is that for many cases it is not clear when a reported serum iron level was drawn with respect to Does the Peak Serum Iron Level Predict the time of ingestion. Therefore, it is difficult to know Toxicity? if a reported level represents a peak level. Significant distribution of iron may have occurred by 6 to 12 The volunteer absorption studies described above hours, so an iron level measured many hours after an suggest that patients with peak serum iron levels ingestion may be well below the peak. between 200 and 300 gg/dL experience gastrointesti- Another problem with serum iron measurement nal symptoms. Among 32 patients with serum iron lev- occurs during deferoxamine therapy. Deferoxamine els between 300 and 500 gg/dL who received various interferes with standard assays for iron and leads to a forms of gastrointestinal decontamination but no falsely lowered serum iron level. 98-1°° This almost cer- chelation therapy, 10 patients were asymptomatic, 18 tainly contributes to nearly universal reports of patients were vomiting, 6 patients had diarrhea, 1 decreased serum iron levels during deferoxamine thera- patient was lethargic, and 1 patient had transient py in clinical case reports. Although accurate serum iron hypotension.97 Although some in this group of patients levels can be determined using atomic absorption spec- probably met our criteria for chelation therapy, all trometry, this is generally not available to the clinician. recovered without sequelae. In another review, 25% of One final problem with the serum iron level is that 13 patients with serum iron levels below 300 gg/dL, unfortunately, even in the ideal case where the exact 64% of 22 patients with levels between 300 and 500 dose of ingested iron is known and we can draw a gg/dL, and 75% of 20 patients with levels greater than level at the exact peak, we almost never can get the 500 gg/dL were moderately or severely poisoned. 95 test result back in a timely fashion. Many physicians Very high serum iron levels, generally greater than do not have access to "stat" serum iron levels. At many 1000 gg/dL, are almost always associated with serious hospitals, the test is run at an outside reference lab or toxicity.7° In the childhood fatalities reported by the batched to be run at regularly scheduled intervals. AAPCC for the past 10 years, 80% of the patients had Generally, the clinician will have to make treatment serum iron levels greater than 1000 pg/dL, and 40% decisions on the basis of the child's clinical status had levels greater than 5000 gg/dL (Table 7). The before a serum iron level is available. reports do not always indicate a time associated with In summary, in spite of factors that may confound the the measured level. interpretation of serum iron levels, these levels gener- Earlier we discussed why the history of ingestion ally correlate with toxicity. Some patients with mild to does not necessarily correlate with the severity of tox- moderately elevated serum iron levels (300-500 pg/dL) icity. Some of the data presented above suggest that are asymptomatic, most have some symptoms, and a the serum iron level may also not correlate with sever- few experience serious toxicity. Patients with levels

Curr Probl Pediatr, March 2000 79 TABLE 8. Utility of the WBC and serum glucose in predicting iron toxicity Parameter Study Sensitivity Specificity PPV NPV WBC >15,000/mm 3 A1°1 0.47 1.0 1.0 0.57 B102 0.12 0.94 0.75 0.42 C1°3 0.28 0.88 0.89 0,24 D±°4 0.15 1.0 1.0 0.22 Glucose >150 mg/dL A 0,38 1.0 1.0 0.48 B 0.07 0.18 0.33 0.41 C 0.25 0.80 0.89 0.24 D 0.09 1,0 1.0 0.21 Vomiting A 0.94 0.25 0.67 0.71 B 0.6 0.71 0.75 0.54 C 0.66 0.32 0.79 0.19 D 0.84 0.5 0.87 0.44 Radiopacities A 0.52 0,81 0.83 0.48 B 0.15 0.94 0,75 0.42 C 0.39 0.66 0.81 0.24 D 0.53 0.5 0.81 0.21 between 500 and 1000 gg/dL experience moderate to The other problem with these criteria is that severe toxicity; however, most recover without seque- although they may identify some of the patients with lae. Patients with levels greater than 1000 pg/dL have elevated serum iron levels and thus exposure, a posi- severe toxicity and are at risk for death, although even tive result from the test does not correlate with a par- in this group, most recover with therapy. 75,76 ticular serum iron level or level of severity. A level of 300 gg/dL is not associated with significant clinical Does the Peripheral Leukocyte Count or toxicity, as discussed above. Serum Glucose Predict Toxicity? Does the Radiograph Predict Toxicity? Early clinical descriptions of iron poisoning demon- A radiograph showing radiopaque pills in the stom- strated that both the peripheral white blood cell count ach or gastrointestinal tract confirms the ingestion (WBC) and the serum glucose were elevated in cases (Fig 4). In many case reports the number of pills seen of serious iron poisoning. 33,95 The utility of these 2 on radiograph has been counted to predict the likeli- parameters as predictors of iron toxicity has been test- hood of toxicity. 105 The accuracy of counting pills on ed in both children and adults. radiograph has never been tested or standardized. One of the first studies in which this question was However, it seems reasonable that the appearance of examined showed that a WBC greater than 15,000 per many pills confirms a significant exposure with the cubic millimeter or a serum glucose greater than 150 potential for toxicity. mg/dL each had 100% specificity and positive predic- However, a "normal" radiograph does not exclude tive value when used to identify patients with serum exposure because not all iron preparations are iron levels greater than 300 gg/dL, lm In other words, equally visible on radiograph. Chewable multivita- the few patients who had an elevated WBC or serum mins with iron can be seen on radiograph but may glucose had elevated serum iron levels and a con- have lower radiopacity than adult iron pills. 1°6 firmed exposure. Unfortunately, the sensitivity of Similarly, the radiopacity of individual iron pills of these tests was only 50%, which means that half of the the same type is not uniform. 1°7,1°8 The degree of patients with a serum iron level greater than 300 gg/dL radiopacity is affected by the number and type of did not have an elevated WBC or serum glucose. pills ingested, the number of pills removed by gas- Several other groups repeated these studies in children trointestinal decontamination, the state of the pills and adults and showed that both the WBC and the and the extent of dissolution after mechanical agita- serum glucose were insensitive as screening tests for tion in the stomach, the location of the pills and elevated serum iron levels (Table 8). l°2-1°4 In general, fragments in the gastrointestinal tract, the degree of vomiting was a more sensitive indicator of elevated absorption, and the formation of any concretions. serum iron levels than either the WBC or serum glu- Besides confirming exposure and estimating the cose, but even vomiting was not 100% sensitive. number of ingested pills, radiographs are also used to

80 Curr Probl Pediatr, March 2000 assess the success of gastrointestinal decontamination. Several reports demonstrated the inability of various methods of gastrointestinal decontamination to remove iron by obtaining radiographs after emesis, lavage, or catharsis. Again, caution must be exercised. When previously seen pills are no longer seen after gastrointestinal decontamination, this may reflect the physical state of the pill or fragments rather than the success of gastrointestinal decontamination. Does the Total Iron-Binding Capacity Predict Toxicity?

For many years it was thought that a total iron- binding capacity (TIBC) greater than the serum iron level was protective against iron toxicity. Unfort- unately the TIBC is not useful in determining the likelihood of toxicity. 109 As an artifact of the techniques used to perform the standard laboratory assay, a rise in the serum iron level FIG 4. Abdominal radiograph. Approximately 40 radiopaque iron leads to an increase in the measured TIBC. Although sulfate pills are visible in the stomach. (From the personal teaching files of the author.) this artifact can be corrected by modifications to the assay, a technician would have to know the serum iron level first to make the appropriate correction. 62,110 both to the urine level of ferrioxamine as well as the However, even the corrected value would not be bet- observer's ability to detect the color change.114 In fact, ter than the clinical examination at deciding the sever- this color change may appear only as darker or brown- ity of poisoning. Deferoxamine also interferes with ish urine, and the best way to note a color change is to the measurement of TIBC. 111 obtain a baseline sample before therapy. It is almost the nature of the change more than the specific color Does the Deferoxamine Challenge Test that is important. Placing sequential urine samples in Predict Toxicity? a test-tube rack is a useful method to evaluate the color changes over time. When deferoxamine is administered parenterally, it Once again a positive challenge will confirm expo- binds Fe ÷3, and the complex, ferrioxamine, is excreted sure, whereas a negative result will not necessarily in the urine. Ferrioxamine is a reddish colored com- "rule out" exposure. In addition, no particular serum pound and imparts to the urine the famous "vin-ros6" iron level or severity of toxicity has a correlation color. with a vin-ros6 urine. 96 Therefore, the deferoxa- In general, significant iron intoxication treated mine challenge test is no longer recommended. Intra- with deferoxamine will lead to a "positive" urine venous deferoxamine is administered therapeutically color change. However, McEneryTM reported that to patients with definite or probable systemic iron poi- only 50% of patients with serum iron levels between soning. 500 and 1000 pg/dL had vin-ros6 urine after defer- A variation of the deferoxamine challenge test has oxamine. Villalobos 112 reported that 9 patients with been proposed to test gastric fluid for iron. In the serum iron levels greater than 350 gg/dL had no deferoxamine "color test," deferoxamine is added to urine color change after deferoxamine therapy. gastric fluid in the presence of H202. A red response Freeman and Manoguerra 113 reported an extraordi- indicates a positive result and the presence of iron as nary case in which a patient had no vin-ros6 color ferrioxamine. Although this test is qualitative and change despite a serum iron level of 1989 pg/dL and not predictive of severity of poisoning, it might con- increased measured urinary iron. firm exposure quickly in a case of unexplained The ability to detect a color change may be related metabolic acidosis.115

Curr Probl Pediatr, March 2000 81 Does Gastrointestinal Decontamination Alter in small children. Iron pills tend to conglomerate and the Course of Iron Poisoninge. congeal, and these sticky masses may also not pass through the lavage tube. The general approach to management of the poi- Although gastric lavage is intended to mechanically soned patient has changed dramatically over the past wash pills out of the stomach, in the case of iron, addi- 25 years. The value of many of the strategies that tional theoretical consideration led to the use of lavage we advocated not so long ago, such as emesis, gas- fluids containing bicarbonate, phosphate, and hydrox- tric lavage, and catharsis, has been questioned. 116 ide ions to complex iron in the gastrointestinal tract to Although each of these therapies was presumed bene- reduce absorption. In the duodenum and small bowel, ficial and was demonstrated (usually in volunteer stud- Fe +2 is absorbed; Fe +3 must be converted to Fe +2 ies) to cause the return of variable amounts of admin- before absorption. Lavage with sodium bicarbonate istered substances and even occasionally to decrease was expected to convert Fe +2 back to Fe +3 and reduce absorption, few studies have shown reduced morbidity absorption. The iron-carbonate complex was also or mortality in actually poisoned patients? 17 In addi- expected to be poorly absorbed. There was evidence tion, there are technical problems associated with per- that nonheme iron was poorly absorbed in the pres- forming emesis or lavage, and each is associated with ence of phosphate and calcium, so lavage with Fleet occasional morbidity or mortality. Phospho-Soda was used as a source of phosphate.119 Pediatricians have been advocates of syrup of In an in vitro model, bicarbonate and phosphate were ipecac; having syrup of ipecac at home has always able to complex a small amount of iron in a simulated been one of the tenets of anticipatory care. However, gastric environment. 12° However, when rats or pigs according to the AAPCC, syrup of ipecac was used in were given 60 mg/kg Fe ÷2 and then lavaged with bicar- only 1.5% of poisoning cases in 1997, compared with bonate, phosphate, or distilled water, absorption of iron 13.4% in 1983. 9 In general, syrup of ipecac is con- was similar in all groups. 72,119 Furthermore, several traindicated in most cases of serious poisoning and is case reports documented hyperphosphatemia after not necessary in the rest. lavage with Fleet's Phospho-Soda? 21'122 As a result of The use of ipecac syrup in the setting of iron poi- the questioned efficacy and demonstrated toxicity, the soning is problematic because it interferes with our use of phosphate lavage was abandoned. Bicarbonate ability to make a clinical assessment. The patient with- lavage was not criticized for toxicity; however, the out vomiting and without other gastrointestinal symp- experience was that there was little clinical efficacy. toms is generally presumed to not be poisoned. Once More recently, magnesium hydroxide was demon- we administer syrup of ipecac and induce emesis, we strated to complex iron in vitro, and it significantly lose one of our most important clinical indicators. reduced the absorption of iron in dogs given 60 mg In the patient with severe iron poisoning who is in Fe+2/kg. 123-125 Magnesium levels were elevated in the shock, syrup of ipecac is contraindicated: the risk of dogs, but there was no clinical toxicity. Magnesium aspiration outweighs the benefit. At least a few of the hydroxide also reduced absorption in human volunteers deaths reported in some of the early cases may have given 5 mg Fe+2/kg.126 Although magnesium hydroxide been caused by aspiration. 31 Most patients with mod- was used as standard therapy in some centers and is erate or severe toxicity will already have vomited by occasionally recommended, gastric lavage with magne- the time of their initial evaluation. Although there sium hydroxide has not gained a strong foothold in the may be some additional recovery of gastric contents, clinical management of iron poisoning?27A28 we do not believe that patients benefit significantly Activated charcoal does not have a major role in the from additional gastric emptying after they have treatment of iron poisoning because metallic ions such vomited spontaneously. 118 as iron do not adsorb to it. However, there may be a Gastric lavage can also be used for gastric emptying. role for activated charcoal in the patient with a mixed In the sick patient, endotracheal intubation can be per- ingestion of a substance such as salicylate, which does formed with a cuffed endotracheal tube to minimize adsorb well to charcoal. Interestingly, activated char- the risks of aspiration. The main problem with gastric coal does bind the deferoxamine-iron complex, so there lavage in the small child is technical; even the largest may be some utility in using oral deferoxamine and orogastric lavage tube may not allow large iron pills to charcoal to decontaminate the gastrointestinal tract or pass easily, and it is difficult to insert the largest tube charcoal hemoperfusion to increase elimination. 129,130

82 Curr Probl Pediatr, March 2000 Whole bowel irrigation has been advocated as the TABLE 9. Indications for deferoxamine therapy ideal procedure for gastrointestinal (as opposed to gas- Shock Altered mental status tric) lavage. This is a technique whereby the entire gas- Persistent gastrointestinal symptoms trointestinal tract is flushed with an isotonic polyelec- Metabolic acidosis trolyte solution. It was originally developed as a means Pills on radiograph of cleansing the gastrointestinal tract before surgery, Serum iron level >500 pg/dL Estimated dose >60 mg Fe+2/kg and it is being used increasingly in poison management. In several animal and volunteer studies in which-whole bowel irrigation was used to treat various experimental agement decisions will be obtained from a thor- poisonings there have been mixed results of effective- ough history and careful physical examination. ness, and none of these studies involved iron. 131 Whole Most of our discussion has been focused on bowel irrigation has been used clinically for iron poi- younger children. However, when managing an soning and is described in a number of case reports and intentional ingestion in a suicidal adolescent, the series. 132-135 In many of the cases, pills were identified clinician should be circumspect about any histori- on abdominal radiographs after emesis and/or lavage, cal information until supportive data are available. and pill fragments were identified in the effluent from For all patients, information about the type and whole bowel irrigation. These patients almost always number of pills ingested and the time of ingestion received parenteral deferoxamine therapy in addition to are important; however this information may be whole bowel irrigation. None of the reported patients inaccurate. The ingestion of almost any amount of had significant complications from the whole bowel iron may lead to gastrointestinal symptoms, and it irrigation, although it is difficult to know whether they is also important to elicit these symptoms in the definitely benefited from the therapy. The American history. It is also important to elicit information Academy of Toxicology concludes that whole bowel about coingested substances, because serious irrigation is an unproved therapy, but many clinical tox- intoxication with other substances may alter the icologists believe that it is the safest method of gas- approach to management. trointestinal decontamination currently available and Subtle presentations require careful attention to that it has the greatest potential for benefit. 131 accurate measurement of vital signs and identifi- A conglomeration of iron pills may form in the cation of early signs of acidosis or shock. Stress, stomach after an overdose. Several case reports have acidosis, gastrointestinal fluid loss from vomiting detailed the use of gastrotomy to remove these iron and diarrhea, third spacing of intravascular vol- bezoars. 136-14° Although patients did well with surgi- ume, and hemoconcentration may lead to lethar- cal therapy, it is unknown whether they would have gy, coma or altered mental status, tachycardia, done as well with traditional therapy alone. Whole tachypnea or hyperpnea, hypotension, pallor, cool bowel irrigation may work well to move such a mass skin, or delayed capillary refill. through the gastrointestinal system. 135,14° . Fluid resuscitation. Most patients with significant intoxication will have some gastrointestinal Clinical Management of Iron Poisoning fluid losses and be in at least the early stages of shock. The initial therapeutic maneuver is to We have discussed at length many of the controver- establish intravenous access and administer nor- sies in the evaluation and management of iron poison- mal (0.9%) saline at an initial dose of 20 mL/kg ing. Using this information we can now construct a followed by a continuous infusion. The total clinical approach. intravenous fluid requirement will be determined 1. Clinical evaluation. Serious iron poisoning is not by the patient's clinical status. difficult to identify: the patient is lethargic, aci- . Laboratory evaluation. The focus is on laboratory dotic, and in shock. A patient in this condition tests that are easily performed and whose results requires immediate attention to airway manage- are available quickly so that they may be used in ment and fluid resuscitation, followed by chela- initial management. tion therapy (see below). A hematocrit or hemoglobin level can be mea- For the patient with a less acute presentation, the sured at the bedside. If there has been significant most important information with respect to man- fluid loss or third-spacing of fluid, the initial

Curr Probl Pediatr, March 2000 83 hematocrit may be elevated. Although significant altered mental status; (3) persistent vomiting, gastrointestinal bleeding can lead to a decreased diarrhea, hematemesis, hematochezia, or other hematocrit, this is not usually seen. A Hemoccult gastrointestinal symptoms; (4) positive anion-gap test of the stool will help to identify occult gas- metabolic acidosis; (5) a large number of pills on trointestinal bleeding. abdominal radiograph; (6) a serum iron level A positive anion-gap metabolic acidosis is seen with greater than 500 gg/dL; or (7) an estimated dose significant iron poisoning and generally identifies a greater than 60 mg Fe+2/kg. patient who will require chelation therapy. An arte- The recommended dose of deferoxamine is 15 rial blood gas and serum electrolytes can quickly mg/kg per hour as a continuous intravenous infu- identify the patient with acidosis and an anion gap. sion. Although this rate of administration is not An electrocardiogram is helpful to evaluate the pos- frequently associated with hypotension, the most sibility of any iron-related myocardial dysfunction common toxicity related to deferoxamine, it is and to screen for other potentially cardiotoxic possible to initiate the infusion at a lower rate, for coingestantS such as cyclic antidepressant agents. example 8 mg/kg per hour, and incrementally An acetaminophen level should be obtained for increase the rate every 5 minutes or so until the all suicidal adolescents or in the setting of an maximum rate is obtained. If intravenous access is unknown ingestion. Acetaminophen poisoning impossible, deferoxamine can be given at a dose may be occult in the early stages but requires ther- of 90 mg/kg intramuscularly. However, intra- apy to prevent toxicity. In a toddler, when the venous fluid therapy is a critical aspect of ingestion has been observed or the child was management, and intravenous access should be found playing with iron pills, screening for aceta- secured as soon as possible. minophen poisoning is not necessary. There are few clear guidelines about the appropri- Clotting abnormalities have been seen in some ate duration of deferoxamine therapy because cases of iron poisoning. Abnormalities of liver there is no well-defined effective dose or end point function will occur with hepatotoxicity, but this is of therapy. The serum iron level is not helpful usually a later manifestation of severe poisoning. because it decreases after 12 to 24 hours naturally A serum iron level will be interesting in retrospect and is falsely lowered in the presence of deferox- but will provide information useful in the early amine. The manufacturer recommends a total dose management only if the result can be available of 6 g, but this would only provide 6 hours of ther- relatively soon after it is drawn. Other tests may apy for an adult at standard infusion rates. be ordered at the discretion of the clinician. It was previously recommended to continue thera- 4. Abdominal radiograph. If the history or physical py until the urine returned to a normal color for 24 examination suggests the possibility of significant hours. This presents the problems of evaluating a expoSure, an abdominal radiograph is useful. A vin-ros6 urine as previously described. In addition, radiograph showing a large number of pills will it leads to protracted courses of therapy that may confirm the exposure and suggest the need for put the patient at increased risk of pulmonary gastrointestinal decontamination and/or treat- toxicity. Chelation therapy should continue until ment. A radiograph negative for pills may be reas- there is significant resolution of systemic toxicity, suring, particularly in the setting of minimal clin- specifically acidosis and shock. Most patients will ical symptoms, but does not eliminate the possi- require chelation for 1 to 2 days, although some bility of exposure. There are no clear guidelines may require longer courses. about how to interpret a radiograph with only a . Gastrointestinal decontamination. Any patient with few pills visible. probable or confirmed significant exposure should 5. Chelation therapy. If the clinical examination, have whole bowel irrigation with the possible laboratory evaluation, or radiograph data suggest exception of a patient who is seen very late (12-24 the possibility of significant exposure or confirm hours) after exposure. Whole bowel irrigation is ini- significant toxicity, deferoxamine therapy should tiated with polyethylene glycol-electrolyte solution be initiated (Table 9). Criteria for initiation of (Colyte, Golytely) at a rate of 0.5 L/h for children or therapy include a history of iron ingestion with (1) 2 L/h for adolescents. Young children will general- any clinical sign of shock; (2) lethargy, coma, or ly require administration of the irrigation solution

84 Curr Probl Pediatr, March 2000 through a nasogastric tube, although adolescents sultation from the toxicologist on-call at the New may choose to drink it. The typical adverse effects York City Poison Control Center by calling 212- of administration are nausea, bloating, vomiting, POISONS (212-340-4494). To find the phone and diarrhea, all of which mimic the symptoms of number or location of a local or regional poison iron poisoning. However, if symptoms seem related center, one can access the Internet home page of to whole bowel irrigation, it is possible to slow the the AAPCC at http://www.aapcc.org. rate of administration to relieve symptoms. Whole bowel irrigation is continued until the effluent is Is There a Modified Approach to clear. An abdominal radiograph may be used to fol- Management in the Pregnant Patient? low the progress of whole bowel irrigation, remem- Prenatal vitamins with iron are prescribed to almost bering the caveats discussed above. all pregnant women, and their widespread availability Although there may be a role for administration of makes iron one of the most common substances ingest- magnesium hydroxide to try to reduce absorption, ed by pregnant women attempting suicide. Maternal there is little clinical experience with its use. There toxicity is generally greater than fetal toxicity. In 2 is no experience to predict the effect of using both reported cases, healthy babies were delivered although whole bowel irrigation and magnesium hydroxide. the mothers died. 141,142 In another case, the mother had 7. Disposition. The patient with no history of abdomi- severe iron toxicity with acidosis, shock, renal failure, nal pain, nausea, vomiting, or diarrhea and with a and disseminated intravascular coagulation but was not normal physical examination is unlikely to have treated with deferoxamine because of concerns about ingested enough iron to lead to serious poisoning. If its teratogenic risks. Instead the mother received an a patient remains asymptomatic for 6 hours, the exchange transfusion that was followed 45 minutes patient may be safely discharged. later by a spontaneous abortion of the 16-week-old Patients with minimal gastrointestinal symptoms fetus. 143'144 Neonatal and cord blood iron levels were only but an otherwise normal physical examina- not elevated. In several cases, pregnant women had tion should have abdominal radiographs, an arteri- signs and symptoms of iron poisoning with elevated al blood gas, and electrolytes. If there are no pills serum iron levels, were treated with deferoxamine, and visible on an abdominal radiograph and there is no delivered healthy infants. 145-151 evidence of metabolic acidosis after several hours, Although the placenta transports iron to the fetus the patient is unlikely to develop systemic toxicity. efficiently, it blocks the transfer of large quantities of A serum iron level less than 500 gg/dL would sup- iron. 3 In a sheep model of iron poisoning, only a small port identifying this as a low-risk patient. This amount of iron was transferred across the placenta patient may be admitted to the hospital for obser- despite significantly elevated serum iron levels. 152 vation or discharged home with close follow-up. Any patient with more than mild gastrointestinal Deferoxamine is reported to be an animal teratogen symptoms or with evidence of altered mental sta- that causes skeletal deformities and abnormalities of tus, shock, or acidosis should receive chelation ossification (Class C pregnancy risk according to the therapy and be admitted to the hospital. Patients FDA). In a recent animal model similar effects were with severe toxicity may require intensive care observed, but only with doses of deferoxamine that therapy. If a hospital has limited pediatric critical caused maternal toxicity. 153 Experimentally, there is care facilities or does not have access to toxicolo- little transfer of deferoxamine across the placenta. 152 gy consultation, plans to transfer a patient to a ter- Therefore, the reported fetal effects may be a result of tiary care hospital should be made early, before the chelation of essential nutrients (such as trace met- the patient becomes unstable for transfer. als) on the maternal side of the placenta. 154 8. Consultation. The regional poison center should In clinical case reports of iron overdose for which be contacted to report any patient who goes to a deferoxamine was used there have been no adverse health care facility for evaluation of possible poi- effects on the fetus, although most have been second soning. For patients who have significant toxicity or third trimester poisonings. 142,145-149,150,151 In 1 case or who are receiving deferoxamine, toxicologic series of 49 patients with iron poisoning during preg- consultation should be requested. If none is avail- nancy, 25 women received deferoxamineJ 55 Only 1 able in your area, you can obtain a telephone con- woman who received deferoxamine in the first

Curr Probl Pediatr, March 2000 85 trimester continued her pregnancy to term; she had a 5. Umbreit JN, Conrad ME, Moore EG, Latour LE Iron healthy infant. One woman who received deferoxa- absorption and cellular transport: the mobilferrirdparaferritin mine at 20 weeks had a premature delivery at 24 weeks, paradim. Semin Hematol 1998;35:12-26. 6. Ryan TP, Aust SD. The role of iron in oxygen-mediated tox- and another woman who received deferoxamine at 30 icities. Crit Rev Toxicol 1992;22:119-41. weeks delivered a baby at term with webbed fingers, a 7. McCord JM. Iron, free radicals, and oxidative injury. Semin malformation believed to be unrelated to the intoxica- Hematol 1998;35:5-12. tion or the therapy. 8. Hershko C, Link G, Ioav C. Pathophysiology of iron over- In support of the safety of deferoxamine in pregnan- load. Ann N Y Acad Sci 1998;850:191-201. cy is the fact that it has been administered without 9. Litovitz TL, Klein-Schwartz W, Dyer KS, Shannon M, Lee S, Powers M. 1997 annual report of the American adverse effects to pregnant women with thalassemia, Association of Poison Control Centers Toxic Exposure as therapy for posttransfusion iron overload. 156-~59 Surveillance System. Am J Emerg Med 1998;16:443-97. Deferoxamine is probably safe for use in pregnant 10. Flomenbaum NE, Howland MAH, Weissman R. Pretty poi- women. Considering the potentially fatal nature of son. Emerg Med 1986;4:69-84. severe iron intoxication, deferoxamine should be 11. Spencer IOB2 Ferrous sulphate poisoning in children. Br Med J 1951;2:112-7. administered to pregnant women with signs and symp- 12. Woolf AD, Lovejoy FH. Epidemiology of drug overdose in toms of significant iron poisoning. children. Drug Saf 1993;9:291-308. 13. Fine JS. Pediatric principles. In: Goldfrank LR, Flomenbaum NE, Lewin NA, Weisman RS, Howland MA, Hoffman RS, Conclusion editors. Goldfrank's toxicologic emergencies. 6th ed. Stamford: Appleton & Lange; 1998. p. 1687-97. In this article we have reviewed the pathophysiology 14. Anonymous. Toddler deaths resulting from ingestion of iron of iron poisoning and suggested an evidence-based supplements-Los Angeles, 1992-1993. MMWR Morb approach to evaluation and management. The frequency Mortal Wkly Rep 1993;42:111-3. of exposure to iron-containing products is high; howev- 15. Anonymous. Preventing iron poisoning in children. FDA er most children do not experience serious toxicity. The backgrounder--current and useful information from the diagnosis of iron poisoning is a clinical one, and ancil- Food & Drug Administration, BG 97-1, amended 1/12/99. Available from: http://www.fda.gov/opacom/backgrounders/ lary tests have poor sensitivity in the identification of ironbg.html. Accessed July 8, 1999. serious toxicity. Most seriously ill patients can be suc- 16. Aldrich RA. Acute iron toxicity. In: Wallerstein RO, Mettier cessfully managed by providing excellent supportive SR, editors. Iron in clinical medicine. Berkely: University of care and chelation therapy with deferoxamine. The California Press; 1958. p. 93-114. small number of severe poisonings may be further 17. Litovitz TL, Clark LR, Soloway, RA. 1993 annual report of the American Association of Poison Control Centers Toxic reduced by decreasing the amount of iron salts available Exposure Surveillance System. Am J Emerg Med 1993; 11: in each pill, packaging high dose iron-containing pills 494-555. as unit-doses, or changing the type of iron products 18. Linakis JG, Frederick KA. Poisoning deaths not reported to available to a less toxic form such as carbonyl iron. the regional poison control center. Ann Emerg Med 1993;22: Future trends in therapy will focus on the use of oral 1822-8. iron chelators as well as new agents being developed to 19. Litovitz T, Manoguerra A. Comparison of pediatric poisoning hazards: an analysis of 3.8 million exposure incidents. limit and treat free radical-induced organ damage. Pediatrics 1992;89:999-1006. 20. Deeths TM, Breeden JT. Poisoning in children--a statistical References study of 1,057 cases. J Pediatr 1971;78:299-305. 21. Fazen LE, Lovejoy FH, Crone RK. Acute poisoning in a 1. Beard JL, Dawson H, Pinero DJ. Iron metabolism: a com- children's hospital: a 2-year experience. Pediatrics 1986;77: prehensive review. Nutr Rev 1996;54:295-317. 144-51. 2. Wells MS, Awad WM. Iron and heme metabolism. In: 22. King WD, Palmisano PA. Ingestion of prescription drugs by Devlin TM, editor. Textbook of biochemistry: with clinical children: an epidemiologic study. South Med J 1989;82: correlations. 3rd ed. New York: Wiley-Liss, Inc; 1992. 1468-71. p.1001-23. 23. Ferguson J, Sellar C, Goldacre MJ. Some epidemiological 3. Andrews NC, Bridges KR. Disorders of metabolism and observations on medicinal and non-medicinal poisoning in sideroblastic anemia. In: Nathan and Oski's hematology of preschool children. J Epidemiol Community Health 1992; infancy and childhood. 5th ed. Philadelphia: WB Saunders 46:207-10. Co; 1998. p. 423-61. 24. Trinkoff AM, Baker SE Poisoning hospitalizations and 4. Harrison PM. Biochemistry of iron. Clin Toxicol 1971; deaths from solids and liquids among children and teenagers. 4:529-44. Am J PuNic Health 1986;76:657-60.

86 Curr Probl Pediatr, March 2000 25. Nightingale SL. From the food and drug administration. of experimental iron poisoning. Ann Emerg Med 1989;18: JAMA 1997;277:1343. 863-6. 26. Huebers HA, Brittenham GM, Csiba E, Finch CA. 45. Artman M, Olson RD, Boucek R J, Boerth RC. Depression Absorption of carbonyl iron. J Lab Clin Med 1986;108: of myocardial contractility in isolated rabbit myocardium 473-8. following exposure to iron: role of free radicals. Toxicol 27. Gordeuk VR, Brittenham GM, Bravo JR, Hughes MA, Appl Pharmacol 1984;72:324-32. Keating LJ. Prevention of iron deficiency with carbonyl iron 46. Link G, Pinson A, Hershko C. Heart cells in culture: a model in female blood donors. Transfusion 1990;30:239-45. of iron overload and chelation. J Lab Clin Med 1985;106: 28. Gordeuk VR, Brittenham GM, McClaren CE, Hughes MA, 147-53. Keating LJ. Carbonyl iron therapy for iron deficiency ane- 47. Tenenbein M, Kopelow ML, deSa DJ. Myocardial failure mia. Blood 1986;67:745-52. and shock in iron poisoning.Human Toxicol 1988;7:281-4. 29. Gordeuk VR, Brittenham GM, Hughes MA, Keating LJ, 48. Link G, Saada A, Pinson A, Konijn AM, Hershko C. Opplt JJ. High dose carbonyl iron for iron deficiency Mitochondrial respiratory enzymes are a major target of iron anemia: a randomized double blind trial. Am J Clin Nutr toxicity in rat heart cells. J Lab Clin Med 1998;131:466-74. 1987;46:1029-34. 49. Robotham JL, Troxler RF, Lietman PS. Iron poisoning: 30. Ling LI, Hornfeldt CS, Winter IR Absorption of iron after another energy crisis. Lancet 1974;2:664-5. experimental overdose of chewable vitamins. Am J Emerg 50. Luongo MA, Bjornson SS. The liver in ferrous sulfate poi- Med 1991;9:24-6. soning: a report of three fatal cases in children and an exper- 31. Thomson JT. Ferrous sulfate poisoning: its incidence, symp- imental study. N Engl J Med 1954;251:995-9. tomatology, treatment, and prevention. Br Med J 1950;1: 51. Gleason WA, deMello DE, deCastro FJ, Connors JJ. Acute 645-6. hepatic failure in severe iron poisoning. J Pediatr 1979;95: 32. Reissman KR, Coleman TJ, Budai BS, Moriarity LR. Acute 138-40. intestinal iron intoxication I: iron absorption, serum iron, 52. Wilzleben CL, Chaffey NJ. Acute ferrous sulfate posioning: and autopsy findings. Blood 1955;10:35-45. a histochemical study of its effect on the liver. Arch Pathol 33. Covey TJ. Ferrous sulfate posioning. Pediatrics 1964; 1966;82:454-61. 64:218-26. 53. Wilzleben CL. An electron microscopic study of ferrous sul- 34. Whitten CF, Gibson GW, Good MH, Goodwin JR Brough fate induced liver damage. Am J Pathol 1966;49:1053-8. AJ. Studies in acute iron poisoning, I: desferioxamine in the 54. Ganote CE, Nahara G. Acute ferrous sulfate hepatotoxicity treatment of acute iron poisoning: clinical observations, in rats: an electron microscopic and biochemical study. Lab experimental studies, and theoretical considerations. Invest 1973;28:426-36. Pediatrics 1965;36:322-35. 55. Wilson SJ, Heath HE, Nelson PL, Ens GG. Blood coagula- 35. Roberts RJ, Nayfield S, Sopcr R, Kent TH. Acute iron intox- tion in acute iron intoxication. Blood 1958;13:483-91. ication with intestinal infarction managed in part by small 56. Eriksson F, Johansson SV, Mellstedt H, Stranberg O, Wester bowel resection. Clin Toxicol 1975;8:3-12. PO. Iron intoxication in two adult patients. Acta Med Scand 36. Tenenbein M, Littrnan C, Stimpson RE. Gastrointestinal 1974; 196:231-6. pathology in adult iron overdose. J Toxicol Clin Toxicol 57. Evensen SA, Forde R, Opedal I, Stormorken H. Acute iron 1990;28:311-20. intoxication with abruptly reduced levels of vitamin 37. Ghandi RK, Robarts FH. Hourglass stricture of the stomach K-dependent coagulation factors. Scand J Haematol 1982; and pyloric stenosis due to ferrous sulphate poisoning. Br J 29:25-30. Surg 1962;49:613-6. 58. Tenenbein M, Israels SJ. Early coagulopathy in severe iron 38. Vuthibhagdee A, Harris NF. Antral stricture as a delayed com- poisoning. J Pediatr 1988;113:697-9. plication of iron intoxication. Pediatric Radiology 1972;103: 59. Tenenbein M. Toxicokinetics and toxicodynamics of iron 163-4. poisoning. Toxicol Lett 1998; 102-3:653-6. 39. Reissman KR, Coleman TJ. Acute intestinal iron intoxica- 60. Bronson WR, Sisson TR. Studies on acute iron poisoning. tion, II: metabolic, respiratory and circulatory effects of Am J Dis Child 1960;99:18-26. absorbed iron salts. Blood 1955;10:46-51. 61. Linakis JG, Lacouture PG, Woolf A. Iron absorption from 40. Whitten CF, Chen YC, Gibson GW. Studies in acute iron chewable vitamins with iron versus iron tablets: implications poisoning, II: further observations on desferrioxamine in the for toxicity. Pediatr Emerg Care 1992;8:321-4. treatment of acute experimental iron poisoning. Pediatrics 62. Burkhart KK, Kulig KW, Hammond KB, Pearson JR, 1966;38:102-10. Ambruso D, Rumack B. The rise in the total iron-binding 41. Whitten CF, Chert YC, Gibson GW. Studies in acute iron capacity after iron overdose. Ann Emerg Med 1991;20:532-5. poisoning, III: the hemodynamic alterations in acute experi- 63. Tenenbein M. Benefits of parenteral deferoxamine for acute mental iron poisoning. Pediatr Res 1968;2:479-85. iron poisoning. J Toxicol Clin Toxicol 1996;34:485-9. 42. Banner W, Tong TG. Iron poisoning. Pediatr Clin North Am 64. Howland MA. Risks of parenteral deferoxamine for acute 1986;33:393-409. iron poisoning. J Toxicol Clin Toxicol 1996;34:491-7. 43. Artman M, Olson RD, Boerth RC. Depression of myocar- 65. Howland MA. Deferoxamine. In: Goldfrank LR, dial contractility in acute iron toxicity in rabbits. Toxicol Flomenbaum NE, Lewin NA, Weisman RS, Howland MA, Appl Pharmacol 1982;66:329-37. Hoffman RS, editors. Goldfrank's toxicologic emergencies. 44. Vernon DD, Banner W, Dean IM. Hemodynamic effects 6th ed. Stamford: Appleton & Lange; 1998. p. 628-32.

Curr Probl Pediatr, March 2000 87 66. Lipschitz DA, Dugard J, Simon MO, Bothwell TH, Charlton 87. Mofenson HC, Caraccio TR, Sharieff N. Iron sepsis: RW. The site of action of desferrioxamine. Br J Haematol Yersinia enterocolitica septicemia possibly caused by an 1971 ;20:395-404. overdose of iron. N Engl J Med 1987;316:1092-3. 67. Moeschlin S, Schnider U. Treatment of primary and sec- 88. Milteer RM, Sarpong S, Poydras U. Yersinia enterocoIitica ondary hemochromatosis and acute iron poisoning with a septicemia after accidental oral iron overdose. Pediatr Infect new, potent iron-eliminating agent (desferrioxamine-B). N Dis J 1989;8:537-8. Engl J Med 1963;269:57-66. 89. Kontoghiorghes GJ. Chemical, pharmacological, toxicolgi- 68. Leikin S, Vossough MD, Mochir-Fatemi E Chelation thera- cal and therapeutic advances of deferiprone (L1) and other py in acute iron poisoning. J Pediatr 1967;71:425-30. iron and aluminum chelators. Arch Toxicol (Suppl) 1996; 69. McEnery JT, Greengard J. Treatment of acute iron ingestion 18:202-14. with deferoxamine in 20 children. J Pediatr 1988;68:773-9. 90. Olivieri NF, Brittenham GM. Long-term trials of deferiprone 70. McEnery JT. Hospital management of acute iron ingestion. in Cooley's anemia. Ann NY Acad Sci 1998;850:217-22. Clin Toxicol 1971;4:603-13. 91. Oliviefi NF, Brittenham GM, McLaren CE, Templeton DM, 71. Westlin WE Deferoxamine as a chelating agent. Clin Cameron RG, McClelland RA, et al. Long-term safety and Toxicol 1971;4:567-602. effectiveness of iron-chelation therapy with deferiprone for 72. Dean B, Oehme FW, Krenzelok E, Hines R. A study of iron thalassemia major. N Engl J Med 1998;339:417-23. complexation in a swine model. Vet Hum Toxicol 1988;30: 92. Kowdley KV, Kaplan MM. Iron-chelation therapy with oral 313-5. deferiprone: toxicity or lack of efficacy? N Engl J Med 73. Jackson TW, Ling LJ, Washington V. The effect of oral 1998;339:468-9. deferoxamine on iron absorption in humans. J Toxicol Clin 93. Fassos FlU, Berkovitch M, Daneman N, Koren L, Cameron Toxicol 1995;33:325-9. RG, Klein J, et al. Efficacy of deferiprone in the treatment of 74. Henretig FM, Karl SR, Weintraub WH. Severe iron poison- acute iron intoxication in rats. J Toxicol Clin Toxicol ing treated with enteral and intravenous deferoxamine. Ann 1996;34:279-87. Emerg Med 1983;12:306-9. 94. Hung O, Manoach S, Howland Marohera B, Goldfrank L, 75. Douglas D, Smilkstein M. Deferoxamine-iron induced pul- Hoffman R. Deferiprone for acute iron posioning [abstract]. monary injury and N-acetylcysteine [abstract]. J Toxicol Clin J Toxicol Clin Toxciol 1997;35:565. Toxicol 1995;33:495. 95. James JA. Acute iron poisoning: assessment of severity and 76. Cheney K, Gumbiner C, Benson B, Tenenbein M. Survival prognosis. J Pediatr 1970;77:117-9. after a severe iron poisoning treated with intermittent infu- 96. Klein-Schwartz W, Oderda GM, Gorman RL, Favin F, Rose sions of deferoxamine. J Toxicol Clin Toxicol 1995 ;33:61-6. SR. Assessment of management guidelines: acute iron poi- 77. Butterworth CE Jr, Scott CW. Iron mobilizing and histamine soning. Clin Pediatr 1990;29:316-21. releasing properties of desferrioxamine [abstract]. Clin Res 97. Bosse GM. Conservative management of patients with mod- 1965;13:33. erately elevated serum iron levels. J Toxicol Clin Toxicol 78. Tenenbein M, Kowalski S, Sienko A, Bowden DH, Adamson 1995;33:135-40. IY. Pulmonary toxic effects of continuous desferoxamine 98. Helfer RE, Rodgerson DO. The effect of deferoxamine on administration in acute iron poisoning. Lancet 1992; the determination of serum iron and iron-binding capacity. J 339:699-701. Pediatr 1966;68:804-6. 79. Anderson KJ, Rivers RPA. Desferrioxamine in acute iron 99. Gevirtz NR, Wasserman LR. The measurement of iron and poisoning. Lancet 1992;339:1602. iron-binding capacity in plasma containing deferoxamine. J 80. Castriota Scanderbeg A, Izzi GC, Bntturini A, Benaglia G. Pediatr 1966;68:802-4. Pulmonary syndrome and intravenous high-dose desferriox- 100. Schenken JR, Grass L. Deferoxamine and antoassay of amine. Lancet 1990;336:1511. serum iron. Clin Toxicol 1971;4:641-2. 81. Freedman MH, Grisaru D, Olivieri N, MacLusky I, Thorner 101. Lacouture PG, Wason S, Temple AR, Wallace DK, Lovejoy PS. Pulmonary syndrome in patients with thalassemia major FH Jr. Emergency assessment of severity in iron overdose by receiving intravenous deferoxamine infusions. Am J Dis clinical and laboratory methods. J Pediatr 1981;99:89-91. Child 1990;14:565-9. 102. Knasel AL, Collins-Barrow MD. Applicability of early indi- 82. Weitman SD, Buchanan GR, Kamen BA. Pulmonary toxici- cators of iron toxicity. J Natl Med Assoc 1986;78:1037-40. ty of deferoxamine in children with advanced cancer. J Natl 103. Chyka PA, Butler AY. Assessment of acute iron poisoning by Cancer Inst 1991;83:1834-5. laboratory and clinical observations. Am J Emerg Med 83. Adamson IYR, Sienko A, Tenenbein M. Pulmonary toxicity 1993;11:99-103. of deferoxamine in iron-poisoned mice. Toxicol Appl 104. Palatnick W, Tenenbein M. Leukocytosis, hyperglycemia, Pharmacol 1993;120:13-9. vomiting, and positive X-rays are not indicators of severity 84. Bentur Y, McGuigan M, Koren G. Deferoxamine (desferriox- of iron overdose in adults. Am J Emerg Med 1996; 14:454-5. amine): new toxicities for an old drug. Drug Saf 1991;6:37-46. 105. Hosking CS. Radiology in the management of acute iron 85. Ward CG, Bullen JJ, Rogers HJ. Iron and infection: new devel- poisoning. Med J Aust 1969; 1:576-9. opments and their implications. J Trauma 1996;41:356-63. 106. Everson GW, Oudjhane K, Young LW, Krenzelok EP. 86. Melby K, Slordahal, Gutterberg TJ, Nordbo SA. Septicemia Effectiveness of abdominal radiographs in visualizing chew- due to Yersinia enterocolitica after oral doses of iron. Br able iron supplements following overdose. Am J Emerg Med Med J 1982;285:487-8. 1989;7:459-63.

88 Curr Probl Pediatr, March 2000 107. Ng RCW, Perry K, Martin DJ. Iron poisoning: assessment of magnesium hydroxide on iron absorption following simulat- radiography in diagnosis and management. Clin Pediatr ed mild iron overdose in human subjects. Acad Emerg Med 1979; 18:614-6. 1998;5:961-5. 108. Staple TW, McAlister WH. Roentgenographic visualization 127. Green VA. Iron ingestions: The Children's Mercy Hospital. of iron preparations in the gastrointestinal tract. Radiology Clin Toxicol 1971;4:245-52. 1964;83:1051-6. 128. Mills KC, Curry SC. Acute iron poisoning. Emerg Med Clin 109. Sift JE, Meldon SW, Tomassoni AJ. Usefulness of the total North Am 1994;12:397-413. iron binding capacity in the evaluation and treatment of 129. Gomez HF, McClafferty HH, Flory D, Brent J, Dart RC. acute iron overdose. Ann Emerg Med 1999;33:73-6. Prevention of gastrointestinal iron absorption by chelation 110. Tenenbein M, Yatschoff RW. The total iron-binding capacity in from an orally administered premixed deferoxamine/ iron poisoning, is it useful? Am J Dis Child 1991;145:437-9. charcoal slurry. Ann Emerg Med 1997;30:587-92. 111. Bentur Y, Klein J, Koren C. Misinterpretation of the iron- 130. Ross MR Banner W, Vernon DD, Pappas J. Charcoal hemo- binding capacity in the presence of deferoxamine. J Pediatr perfusion in experimental iron intoxication [abstract]. Vet 1991;118:139-42. Hum Toxicol 1990;32:349. 112. Villalobos D. Reliability of urine color changes after defer- 131. Tenenbein M. Position statement: whole bowel irrigation. J oxamine challenge [abstract]. Vet Hum Toxicol 1992; 34: 330. Toxicol Clin Toxicol 1997;35:753-62. 113. Freeman DA, Manoguerra AS. Absence of urinary color 132. Tenenbein M. Whole bowel irrigation in iron poisoning. J change in a severely iron poisoned child treated with defer- Pediatr 1987; 111:142-5. oxamine [abstract]. Vet Hum Toxicol 1981 ;23:351. 133. Palatnick W, Tenenbein M. Safety of treating poisoning 114. Eisen TF, Lacouture PG, Woolf A. Visual detection of fer- patients with whole bowel irrigation. Am J Emerg Med 1988; rioxamine color changes in urine. Vet Hum Toxicol 1988; 6:200-1. 30:369-70. 134. Everson GW, Bertaccini EJ, O'Leary J. Use of whole bowel 115. McGuigan MA, Lovejoy FH, Marino SK, Propper RD, irrigation in an infant following iron overdose. Am J Emerg Goldman R Qualitative deferoxamine color test for iron Med 1991;9:366-9. ingestion. J Pediatr 1979;94:940-2. 135. Kaczorowski JM, Wax PM. Five days of whole-bowel irri- 116. Smilkstein MJ. Techniques used to prevent gastrointestinal gation in a case of pediatric iron ingestion. Ann Emerg Med absorption of toxic compounds. In: Goldfrank LR, 1996:27:258-63. Flomenbaum NE, Lewin NA, Weisman RS, Howland MA, 136. Peterson CD, Fifield CS. Emergency gastrotomy for acute Hoffman RS, editors. Goldfrank's toxicologic emergencies. iron poisoning. Ann Emerg Med 1980;9:262-4. 6th ed. Stamford: Appleton & Lange; 1998. p. 35-51. 137. Venturelli J, Kwee Y, Morris N, Cameron G. Gastrotomy in 117. Kulig K, Bar-Or D, Cantrill SV, Rosen R Rumack BH. the management of acute iron poisoning. J Pediatr 1982;100: Management of acutely poisoned patients without gastric 768-9. emptying. Ann Emerg Med 1985;14:562-7. 138. Foxford R, Goldfrank L. Gastrotomy: a surgical approach to 118. Boxer L, Anderson FP, Rowe DS. Comparison of ipecac- iron overdose. Ann Emerg Med 1985; 14:1223-6. induced emesis with gastric lavage in the treatment of acute 139. Landsman I, Bricker JT, Reid BS, Bloss RS. Emergency salicylate ingestion. J Pediatr 1969;74:800-3. gastrotomy: treatment of choice for iron bezoar. J Pediatr 119. Dean BS, Krenzelok ER In vivo effectiveness of oral com- Surg 1987;22:184-5. plexation agents in the management of iron poisoning. J 140. Tenenbein M, Wiseman N, Yatscoff RW. Gastrotomy and ToxicoI Clin Toxicol 1987;25:221-30. whole bowel irrigation in iron poisoning. Pediatr Emerg 120. Czajka PA, Konrad JD, Duffy JR Iron poisoning: an in vitro Care 1991;7:286-8. comparison of bicarbonate and phosphate lavage solutions. J 141. Richards R, Brooks SEH. Ferrous sulphate poisoning in Pediatr 1981;98:491-4. pregnancy with afibrogenaemia as a complication. West 121. Bachrach L, Correa A, Levin R, Grossman M. Iron poison- Indian Med J 1966;15:134-40. ing: complications of hypertonic phosphate lavage therapy. J 142. Olenmark M, Biber B, Dottori O, Rybo G. Fatal iron intoxi- Pediatr 1979;94:147-9. cation in late pregnancy. J Toxicol Clin Toxicol 1987; 122. Geffner ME, Opas LM. Phosphate poisoning complicat- 25:347-59. ing treatment for iron ingestion. Am J Dis Child 1980; 143. Manoguerra AS. h'on poisoning: report of a fatal case in an 134:509-10. adult. Am J Hosp Pharm 1976;33:1088-90. 123. Decker WJ, Combs HF, Corby DG. Adsorption of drugs and 144. Strom RL, Schiller R Seeds AF, ten Bensel R. Fatal poison- poisons by activated charcoal. Toxicol Appl Pharmacol 1968; ing in a pregnant female. Minn Med 1976;59:483-9. 13:454-60. 145. Blanc R Hryhorczuk D, Danel I. Deferoxamine treatment of 124. Decker WJ, Shafik HM, Corby DG, Milburn RE. acute iron intoxication in pregnancy. Obstet Gynecol Sequestration of iron by magnesium oxide: an in vitro study. 1984;64 (3 Suppl):12S-4. Vet Hum Toxicol 1981;23 (Suppl 1):33-4. 146. Rayburn WF, Donn SM, Wulf ME. Iron overdose during 125. Corby DG, McMullen AH, Chadwick EW, Decker WJ. pregnancy: successful therapy with deferoxamine. Am J Effect of orally administered magnesium hydroxide in Obstet Gynecol 1983;147:717-8. experimental iron intoxication. J Toxicol Clin Toxicol 1985; 147. Turk J, Aks S, Ampuero F, Hryhorczuk D. Successful therapy 23:489-99. of iron intoxication in pregnancy with intravenous deferoxam- 126. Wallace KL, Curry SC, LoVecchio F, Raschke RA. Effect of ine and whole bowel irrigation. Vet Hum Toxicol 1993 ;35:441-4.

Curr Probl Pediatr, March 2000 89 148. Khoury S, Odeh M, Oettinger M. Deferoxamine treatment Matemal-feta! toxicology: a clinician's guide. New York for acute iron intoxication in pregnancy. Acta Obstet (NY): Marcel Dekker; 1990. p. 89-114. Gynecol Scand 1995;74:756-7. 155. McElhatton PR, Roberts JC, Sullivan FM. The consequences 149. Lacoste H, Goyert GL, Goldman LS, et al. Acute iron intox- of iron overdose and its treatment with desferrioxamine in ication in pregnancy. Obstet Gynecol 1992;80:500-1. pregnancy. Hum Exp Toxicol 1991;10:251-9. 150. Schauben JL, Augenstein WL, Cox J, Sato R. Iron posion- 156. Thomas RM, Skalicka AB. Successful pregnancy in trans- ing: report of three cases and a review of therapeutic inter- fusion-dependent thalassemia. Arch Dis Child 1980;55: vention. J Emerg Med 1990;8:309-19. 572-4. 151. Tran T, Wax JR, Steinfeld JD, Ingardia CJ. Acute intentional 157. Martin K. Successful pregnancy in b-thalassemia major. iron overdose in pregnancy. Obstet Gynecol 1998;92:678-80. Aust Paediatr J 1983;19:182-3. 152. Ctm2¢ S, Bond GR, Rashke R, et al. An ovine model of mater- 158. Christiaens GCML, Rijksen G, Marx J, et al. Desferrioxamine nal iron poisoning in pregnancy.Ann Emerg Med 1990;19:632-8. in pregnancy [abstract]. Arch Gynecol 1985;237(suppl):80. 153. Bosque MA, Domingo JL, Corbella J. Assessment of the 159. Voskaridou E, Konstantopoulous K, Kyriakou D, developmental toxicity of deferoxamine in mice. Arch Loukopoulous D. Deferoxamine treatment during early preg- Toxicol 1995;69:467-71. nancy: absence of teratogenicity in two cases. Haematologica 154. Tenenbein, M. Poisoning in pregnancy. In: Koren G, editor. 1993;78:183-4.

90 Curr Probl Pediatr, March 2000