Pharmacobezoars Described and Demystified

Clinical Toxicology

ISSN: 1556-3650 (Print) 1556-9519 (Online) Journal homepage: https://www.tandfonline.com/loi/ictx20

Pharmacobezoars described and demystified

Serge-Emile Simpson

To cite this article: Serge-Emile Simpson (2011) Pharmacobezoars described and demystified, Clinical Toxicology, 49:2, 72-89, DOI: 10.3109/15563650.2011.559472 To link to this article: https://doi.org/10.3109/15563650.2011.559472

Published online: 03 Mar 2011.

Submit your article to this journal

Article views: 641

Citing articles: 25 View citing articles

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=ictx20 Clinical Toxicology (2011) 49, 72–89 Ó 2011 Informa Healthcare USA, Inc. ISSN 1556-3650 print/ISSN 1556-9519 online DOI: 10.3109/15563650.2011.559472

REVIEW Pharmacobezoars described and demystiﬁed

SERGE-EMILE SIMPSON

Albert Einstein Medical Center, Department of Emergency Medicine, Philadelphia 19141, USA

Introduction. A bezoar is a concretion of foreign material that forms and persists in the gastrointestinal tract. Bezoars are classified by their material origins. Phytobezoars contain plant material, trichobezoars contain hair, lactobezoars contain milk proteins, and pharmacobezoars contain pharmaceutical products. Tablets, suspensions, and even insoluble drug delivery vehicles can, on rare occasions, and sometimes under specific circumstances, form pharmacobezoars. The goal of this review is to catalog and examine all of the available reports in the English language medical literature that convincingly describe the formation and management of pharmacobezoars. Methods. Articles included in this review were identified by performing searches using the terms ‘‘bezoar,’’ ‘‘pharmacobezoar,’’ and ‘‘concretion’’ in the following databases: OVID MEDLINE, PubMed, and JSTOR. The complete MEDLINE and JSTOR holdings were included in the search without date ranges. The results were limited to English language publications. Articles that described nonmedication bezoars were not included in the review. Articles describing phytobezoars, food bezoars, fecal impactions, illicit drug packet ingestions, enteral feeding material bezoars, and hygroscopic diet aid bezoars were excluded. The bibliographic references within the articles already accumulated were then examined in order to gather additional pharmacobezoar cases. The cases are grouped by pharmaceutical agent that formed the bezoar, and groupings are arranged in alphabetical order. Discussions and conclusions specific to each pharmaceutical agent are included in that agent’s subheading. Discussion. Patterns and themes that emerged in the review of the assembled case reports are reviewed and presented in a more concise format. Conclusion. Pharmacobezoars form under a wide variety of circumstances and in a wide variety of patients. They are difficult to diagnose reliably. Rules for suspecting, diagnosing, and properly managing a pharmacobezoar are highly dependent on the pharmaceutical agent or agents involved. Becoming familiar with the sparse data available on pharmacobezoars and maintaining a high index of suspicion in future clinical encounters may be the best way to improve diagnostic sensitivity and accuracy.

Keywords Gut and hepatotoxicity; Organ/tissue speciﬁc; Complications of poisoning; GI; Pharmaceuticals; Other; Gastrointestinal decontamination

Introduction under specific circumstances, form pharmacobezoars. The goal of this review is to catalog and examine all of the A bezoar is a concretion of foreign material that forms and available reports in the English language medical literature persists in the gastrointestinal (GI) tract. The word that convincingly describe the formation and management ‘‘bezoar’’ is thought to be derived from the Arabic word of pharmacobezoars. ‘‘badzehr’’ or the Persian word ‘‘panzehr,’’ both of which are terms meaning counterpoison.1 Hard, stone-like bezoars Methods occasionally discovered in a stomach of a ruminant animal, such as a cow or a goat, were prized in antiquity as antidotal Articles included in this review were identified by perform- talismans. Bezoars found in humans are classified by their ing searches using the terms ‘‘bezoar,’’ ‘‘pharmacobezoar,’’ material origins. Phytobezoars contain plant material, and ‘‘concretion’’ in the following databases: OVID trichobezoars contain hair, lactobezoars contain milk MEDLINE, PubMed, and JSTOR. The complete MEDLINE proteins, and pharmacobezoars contain pharmaceutical and JSTOR holdings were included in the search without products. Tablets, suspensions, and even insoluble drug date ranges. The results were limited to English language delivery vehicles can, on rare occasions, and sometimes publications. Resulted articles were then screened for appropriate content. Articles that described nonmedication bezoars were not included in the review. Articles describing phytobezoars, food bezoars, or fecal impactions that may Received 1 December 2010; accepted 28 January 2011. have formed under the indirect influence of a drug were Address correspondence to Serge-Emile Simpson, MD, Albert excluded. Articles describing illicit drug packet ingestions Einstein Medical Center, Department of Emergency Medicine, were excluded. Articles describing pure enteral feeding 5501 Old York Road, Philadelphia 19141, USA. E-mail: material bezoars were not included, as these are food [email protected] bezoars despite the fact that feeds may be pharmaceutical

72 Pharmacobezoars described and demystified 73 products. Similarly, articles describing bezoars caused by noted. On laparotomy, the authors discovered ‘‘a large bolus hygroscopic diet aids, such as glucomannan or psyllium of inspissated charcoal in the caecum’’ and evidence of local fiber, were excluded because they are not strictly therapeutic bowel ischemia without perforation. The patient underwent products. Published opinions or letters to the editor that a right hemicolectomy and survived. The authors concluded discussed pharmacobezoars and their management were that the patient’s critical illness and the lack of cathartic included in this review. The articles collected in the manner coadministration with MDAC were causative factors in the described above, however, proved to be an incomplete formation of the bezoar. catalog of eligible literature. The bibliographic references Merriman and Stokes5 reported an unusual case of AC within the articles already accumulated were then examined bezoar-induced SBO. The patient was a teenaged boy with a in order to gather additional pharmacobezoar cases. history of a high gastric reduction 7 months earlier, anorexia nervosa, and severe malnutrition who was being treated for a Activated charcoal combined opiate, benzodiazepine, and tricyclic antidepres- There are a number of case reports that suggest therapeutic sant (TCA) overdose. He was given two 50 g doses of AC, 4 h use of activated charcoal (AC) can generate hard concretions apart, without a cathartic. Within 24 h, the patient developed a in the GI tract. AC bezoars typically cause a mechanical GI bowel obstruction and underwent laparotomy. A 6 cm 6 5cm tract obstruction without systemic toxicity. charcoal bezoar was identified in the mid-jejunum and Watson et al.2 reported an unconfirmed case of intestinal manually fragmented without the need for enterotomy or obstruction caused by AC. The patient was an adult man bowel resection. In comparison to the cases above, this patient being treated for a serious acute overdose of carbamazepine received a small amount of charcoal. The authors suggested and thioridazine. The patient was intubated and started on a that the combined anticholinergic and opiate ingestion, multidose activated charcoal (MDAC) regimen (240 g in the together with the patient’s history of bariatric surgery, perhaps first 24 h) along with repeat doses of magnesium citrate via slowed the transit of the charcoal. nasogastric (NG) tube. Clinical and radiographic signs of The capacity for inspissated AC to cause perforation was bowel obstruction developed within 24 h. Magnesium demonstrated in a case report by Gomez et al.6 The patient citrate cathartics as well as multiple saline enemas was a young woman who underwent MDAC therapy for an encouraged the patient to pass several ‘‘hard, black chunks’’ apparently mild amitriptyline overdose. The patient received of particulate matter per rectum, and the obstruction was two 50 g of AC over 4 h, and developed peritonitis 2 days resolved. MDAC was resumed without further incident. The later. Her surgeons discovered an obstructing AC bezoar authors did not actually demonstrate an AC bezoar at the site and a 4-cm diameter perforation together at the sigmoid where it formed. colon. The patient’s methadone use and hospital prescribed Ray et al.3 demonstrated an AC bezoar during lapar- benztropine may have contributed to the obstruction. The otomy. Their patient was a young man being treated for a authors noted that cathartics were not used. In another case massive amitriptyline overdose. The patient was critically report, by Mizutani et al.,7 a patient receiving MDAC for an ill, requiring blood pressure support with epinephrine and organophosphate insecticide ingestion developed a serious military antishock trousers. He received phenytoin, diaze- hemorrhagic rectal ulcer after passing several ‘‘coarse- pam, and morphine for treatment of seizures and agitation. surfaced briquettes’’ measuring 3–5 cm in diameter in her The authors administered 30–60 g of AC every 4–6 h for 5 stool. This occurred at least 7 days after a 50-h course of days. It is unclear whether a cathartic was coadministered or MDAC was completed, despite the use of magnesium salt not. On day 3, the patient’s abdomen had become distended. cathartics and continuous dinoprost trometamol injections On day 5, the patient underwent laparotomy for clinical intended to prevent constipation. Contributing factors in this signs of a small bowel obstruction (SBO), whereupon a case may have been the atropine used to treat the patient’s 3cm6 3 cm hard black mass was observed in the distal prolonged cholinergic toxidrome and the large amount of small bowel. The authors state that the mass was inspissated charcoal ingested (between 400 and 625 g). charcoal but did not specify how the obstruction was Goulbourne and Cisek8 reported a case wherein a 64- relieved. The authors speculated that the anticholinergic year-old woman with a history of chronic bronchitis and properties of amitriptyline, the constipating effect of prior hysterectomy was started on MDAC to treat a chronic morphine, and the patient’s profound hypotension all theophylline overdose. In this case, there was no history of contributed to bezoar formation. massive ingestion or deviation from her prescribed medica- The next case of confirmed AC bezoar was reported by tion regimen, and there is no stated history of an Atkinson et al.4 The patient in this case overdosed on a anticholinergic or opiate coingestion. The patient received barbiturate and two benzodiazepines. The patient also 350 g of AC over approximately 24 h and 96 g of sorbitol suffered from extreme exposure and was initially hypother- with her first dose of AC. She was discharged home after mic, hypotensive, and obtunded. The patient received 25 g having a charcoal-laden bowel movement but returned 5 of AC every 2 h with 200 mL of water via NG tube. No days later with a SBO. At laparotomy, it was noted that the cathartic was given. After 18 h the patient remained patient had ‘‘low-grade adhesions’’ and an AC bezoar at the constipated, and the MDAC therapy was discontinued. At ileocecal valve. It is reasonable to conclude, in this case, that 48 h, clinical and radiographic signs of obstruction were the patient’s adhesions may have contributed to the

Clinical Toxicology vol. 49 no. 2 2011 74 S.-E. Simpson formation of an obstructing AC bezoar. An interesting formation.4–6 Factors that may predispose a patient to an ileus feature of this case is the delayed presentation of the bezoar. are frequently present. Critical illness, hypotension, hypother- A case of a delayed large bowel obstruction caused by an mia, opiate drug effects, anticholinergic drug effects, and AC bezoar appears in a review article by Mauro et al.9 That preexisting intestinal adhesions may all contribute to the case, originally reported by Flores and Battle,10 is no longer formation of AC bezoars. AC bezoars may be associated with indexed in the PubMed or Medline databases, and therefore, bowel perforation and GI mucosal injury. Treatment of AC details are unavailable. More recently, Chan et al.11 report a bezoars depends on the severity and location of the obstruction. case of delayed onset obstructing AC bezoar. Enemas may be effective for colonic bezoars, but surgical Dorrington et al.12 reported the results of a retrospective, interventions are typically needed to relieve small bowel multicenter chart review that attempted to estimate the bezoars. The role of cathartics or other dilutional strategies in frequency of pulmonary and GI complications associated preventing the occurrence of AC bezoars is controversial. In all with MDAC therapy in their geographic region. More than but one of the cases, cathartics were not given to the patient 6000 charts were reviewed from eight tertiary care hospitals beyond a single dose. Several of the authors of these papers and three pediatric hospitals in Western Canada and Oregon. suggest that inadequate cathartic administration contributed to The authors identified 878 patients in a 5-year span who bezoar formation in their patients. The exception is the case received MDAC but concluded that a clinically significant reportedbyWatsonetal.,2 wherein repeat doses of magnesium GI obstruction occurred in none of those patients. This study citrate failed to prevent an obstruction, yet cathartic enemas underscores the rarity of obstructing charcoal bezoars. apparently relieved one. Since MDAC therapy by definition There are no reports of single doses of AC forming requires frequent dosing, coadministration of a cathartic can bezoars. With the possible exception of the Merriman and expose patients to the known risks of cathartic-induced fluid Stokes’13 case, confirmed charcoal bezoars in humans have and electrolyte imbalances. Therefore, the use of osmotically resulted from MDAC therapy. MDAC therapy is defined in active cathartic agents such as sorbitol and magnesium salts the American Academy of Clinical Toxicology and cannot be recommended as an adjunct to MDAC. It is likely, European Association of Poisons Centres and Clinical though, that a diluting strategy can help prevent the inspissation Toxicologists (AACT/EAPCCT) position statement on of ingested AC. Polyethylene glycol electrolyte lavage solution MDAC as more than two doses of AC administered to a (PEG-ELS) is a nonabsorbable, isotonic oral lavage solution patient in order to enhance the elimination of a systemically commonly used as a bowel irrigant. PEG-ELS is a safer absorbed drug. The authors of the position paper advised alternative to osmotically active cathartics. Currently, there are that MDAC be considered for the treatment of life- no publications that have commented on the safety and efficacy threatening carbamazepine, dapsone, phenobarbital, quinine, of using PEG-ELS as an adjunct to MDAC therapy. A dosing or theophylline overdoses, especially when the use of strategy used by this author is 250 mL of PEG-ELS delivered MDAC can spare the patient the need for more invasive slowly via NG tube following each dose of AC. procedures (e.g. charcoal hemoperfusion). TCA overdose is excluded from the list of indications. The paper proposed Aluminum hydroxide that an intestinal obstruction and an anatomically nonintact Aluminum hydroxide (Al(OH)3) bezoars have been de- GI tract should be absolute contraindications and that scribed in the medical literature on many occasions. opiate- or anticholinergic-induced GI tract dysmotility be Al(OH)3 has minor antacid properties and phosphate- relative contraindications for MDAC therapy. The authors binding properties. As such, Al(OH)3 preparations are stated that optimal dosing for MDAC is unknown but popular over-the-counter treatments for dyspepsia and recommend an adult starting a dose of 50–100 g followed commonly prescribed to patients with chronic renal failure by additional doses at a minimum rate of 12.5 g/h. The to prevent hyperphosphatemia. In the era before widespread AAPCC/EAPCC authors advised against the use of availability of intravenous (IV) antacid medications, coadministered cathartics, citing their unproven value and Al(OH)3 suspensions were also used to treat GI hemorrhage the risk of iatrogenic fluid and electrolyte shifts in children. (GIH). Al(OH)3 is a constipating medication, and the This recommendation diverges strongly from the recom- distinction between a fecal impaction induced by Al(OH)3 9 mendations of Mauro et al. These authors in looking at therapy and an actual Al(OH)3 concretion can be difficult to eight cases of AC-related constipation and obstruction, make. Therefore, the discussion here will be limited to concluded that ‘‘activated charcoal therapy should always gastric or small bowel Al(OH)3 bezoars. include a cathartic which is given in intermittent doses as The first reported case of Al(OH)3 bezoar dates from determined by the patient’s clinical condition.’’ 1939.14 A 59-year-old woman developed a bezoar while In conclusion, AC bezoars are very rare and occur almost receiving a continuous infusion of 2 L of amphogel per day exclusively as a complication of MDAC therapy. Bowel via NG tube to treat a serious GIH. Adjunctive therapies obstruction symptoms can appear within 24 h or can be delayed included gelatin water, corn syrup, oral liquid petrolatum, for several days, depending on how proximal or distal the site of and morphine sulfate. After 7 days of this therapy, the obstruction is. The majority of bezoars occur in patients patient developed the clinical signs of a SBO and expired receiving more than 350 g of AC in less than 48 h. However, within hours. At autopsy, a firm, putty-like material mixed there does not seem to be a minimum AC dose for bezoar with blood was found densely obstructing the distal half of

Clinical Toxicology vol. 49 no. 2 2011 Pharmacobezoars described and demystified 75 her ileum and her proximal cecum. No evidence of In summary, Al(OH)3 is known to be constipating under preexisting intestinal obstruction or stricture was found. normal circumstances and is known to form insoluble The author suggested that the patient’s critical illness, as aluminum phosphate precipitates in the GI tract. When well as her exposure to morphine and gelatin, encouraged ingested chronically in large doses, Al(OH)3 has the the formation of the bezoar. Similar cases followed. potential to turn into an obstructing sludge. Obstruction Townsend et al.15 described three critically ill patients with may result from the formation of a pharmacobezoar or a renal failure and GIH, all of whom developed bowel fecal impaction. Critically ill patients are at particular risk, obstructions requiring surgical management. In two of these and agents that decrease bowel motility or gastric juice patients, 30–60 mL of Al(OH)3 was administered hourly for secretion may contribute to bezoar formation. Patients tend about 5 days, and each developed a SBO. Descriptions of to present with symptoms and radiographic evidence of the obstructing material matched the description in the 1939 acute bowel obstruction. The bezoar itself appears to be case. Sodium polystyrene sulfonate may have contributed to radiolucent, however. Conservative measures, such as NG bezoar formation in one of Townsend’s patients. Korenman suction, cathartic administration, and enemas are not likely 16 et al. described two patients who expired after each to be successful in dislodging a large Al(OH)3 bezoar, and developing a SBO while taking Al(OH)3 gel. One of these surgical decompression should be considered early. Out- patients was being treated for a massive burn complicated comes in critically ill patients who develop Al(OH)3 bezoars by a GIH. He or she developed an obstructing bezoar while are typically poor, whereas healthier individuals seem to receiving 400–600 mL of Al(OH)3 daily over the course of 8 recover after removal of the obstructing bezoar. The case days. The second patient had chronic renal insufficiency and reports available did not address the controversial matter of developed his or her bezoar while taking 120 mL of potential aluminum toxicity in these patients. Al(OH)3 daily to treat hyperphosphatemia. This patient had Aspirin, enteric coated a colonic perforation at laparotomy and inspissated Al(OH)3 throughout his or her small and large bowels. Girotti et al.17 Enteric-coated aspirin (ECASA) is a form of delayed release coined the term ‘‘amphojeloma’’ to describe the massive aspirin tablet. ECASA tablets feature a coating designed to Al(OH)3 bezoar found obstructing the small bowel of their prevent the dissolution of the tablet in the acidic environment critically ill patient. Unfortunately, despite aggressive of stomach. The coating looses integrity at the relatively surgical decompression of the obstruction, their patient higher pH environment of the duodenum and small bowel. eventually died of overwhelming sepsis. In these assembled The first ECASA product marketed to the public was cases, obstructing bezoars are seen forming in critically ill EcotrinTM, developed by Smith, Kline, and French in the 21 TM patients ingesting large amounts of Al(OH)3 over time. 1950s. Ecotrin employed a cellulose-acetate-phthalate A few Al(OH)3 bezoar case reports break from the pattern (CAP) coating, which remained stable at pH 5 6.0–6.8. described above. Potyk18 described a unique case of Today, CAP is a common pharmaceutical excipient. obstruction in a 26-year-old man caused by antacid tablets. ECASA bezoars have been described on multiple 22 The tablets contained Al(OH)3 and magnesium hydroxide and occasions in the medical literature. Harris described a were designed to be chewed before swallowing. The patient gastric ECASA bezoar in a 68-year-old woman with a had instead been swallowing the hard, insoluble tablets whole history of rheumatoid arthritis (RA) and occult hypertrophic for several months. Plain radiography of the patient’s pyloric stenosis. The patient had been on a stable dose of abdomen revealed ‘‘opaque shadows’’ resembling tablets. ECASA for 11 years but developed relapsing anorexia, At laparotomy, 150 intact tablets were found impacted at the weight loss, and epigastric pain 2–3 months prior to site of an ileal stricture. The patient did well after the presentation. Oral contrast-enhanced radiography revealed obstruction was surgically relieved. Burruss et al.19 described a large pill mass and a gastric outlet obstruction (GOO). a 63-year-old man without preexisting GI pathology who Subsequent gastrotomy revealed a slurry containing 61 developed a SBO after taking 300 mL of MaaloxTM and intact ECASA tablets and remnant shells of dissolved 300 mg of oral ranitidine daily for 3 weeks. The surgeons in ECASA tablets within a dilated stomach. Harris’ patient did this case sent the 3 cm 6 5 cm chalky mass for chemical not exhibit the classic signs of salicylate toxicity, and her analysis. The mass contained aluminum and magnesium, serum salicylate concentration (SSC) at presentation was which was consistent with the ingredients in MaaloxTM. 2.6 mg/dL. Harris also reported that the 61 tablets recovered Burruss et al. suggested that the ranitidine may have from his patient were later dissolved within 30 min in a contributed to the formation of the bezoar by reducing the solution of pH ¼ 6.8, thus proving that the coating was volume of the patient’s gastric secretions. These two cases are manufactured properly. notable in that they involve noncritically ill patients and over- Sogge et al.23 described a second case of an apparent the-counter products. A third case, reported by Portuguez- ECASA bezoar. This case again involved in patient on a Malavasi and Aranda,20 describes a full-term neonate who stable daily dose of ECASA who came to medical attention developed what appeared on oral contrast-enhanced radio- with symptoms not associated with salicylate toxicity. A graphs to be a gastric amphojeloma. The bezoar was not barium sulfate (BaSO4)-enhanced plain radiograph demon- confirmed by endoscopy and appeared to resolve with regular strated both delayed gastric emptying and a large collection saline irrigations of the stomach via NG tube. of tablets in the stomach. Abdominal massage and gastric

Clinical Toxicology vol. 49 no. 2 2011 76 S.-E. Simpson lavage (GL) were unsuccessful at dislodging the tablets. The the authors report a posthemodialysis SSC of 46 mg/dL. The authors then successfully cleared the bezoar with a novel patient required repeat hemodialysis for ‘‘persistent symp- technique of alternating NG infusions of a sodium toms’’ and SSC that rose to 73 mg/dL. Two weeks after this bicarbonate (NaHCO3) solution and NG aspirations every presentation, the patient underwent gastric surgery, where- 30 minutes over 24 h. Presumably this technique raised the upon the GOO was noted as was the absence of any intraluminal pH of the stomach above 6.0, dissolved the medication tablets. The authors postulated that the patient enteric coating on the tablets, and then aspirated the aspirin was experiencing episodes of acute-on-chronic salicylism. slurry. The patient’s SSC on presentation was 0.0 mg/dL. They suggested that the patient’s history of prior vagotomy Twelve hours after the start of the bicarbonate lavage, the and occasional antacid use may periodically have increased level was 22.5 mg/dL, and by 24 h, it was 10 mg/dL. It is her gastric pH above 6.0 and allowed collected ECASA unclear whether the patient ever developed symptoms of tablets in her stomach to dissolve en masse. salicylate toxicity. Bogacz and Caldron27 reported an interesting case wherein Sherry described a 95-year-old woman with a history of a 77-year-old woman with RA and occult GOO was admitted duodenal ulcer who required gastrotomy to remove 38 intact for early satiety, nausea, anorexia, and weight loss. BaSO4- ECASA tablets from her stomach.24 Again, the pills were enhanced plain radiography revealed multiple tablets in her taken under normal therapeutic conditions, and the patient stomach. The patient’s initial SSC was 53.0 mg/dL. After had significant peptic ulcer disease that likely impaired the BaSO4 study, the patient’s SSC increased to 31.4 mg/dL. gastric emptying. BaSO4-enhanced radiographs identified GL with NaHCO3 solution resulted in partial clearance of the the bezoar initially. The patient had a mildly elevated SSC bezoar. Endoscopy revealed additional retained ECASA of 2.8 mmol/L (*38.7 mg/dL) on presentation. tablets and a prepyloric ulcer. The authors then employed the 25 Springer and Groll presented a case of an elderly NaHCO3 infusion/aspiration procedure initially described by woman with occult pyloric stenosis who started taking Sogge et al.23 to clear the remainder of the bezoar. Bogacz ECASA several months before presentation. The patient and Caldron suspected that perhaps the BaSO4 may have developed some abdominal discomfort, distention, nausea, partially dissolved the tablets. They performed an in vitro and mild tinnitus for 1 week. An outpatient BaSO4 meal experiment with two brands of BaSO4, with and without the study revealed an ECASA tablet mass in her stomach. No addition of effervescent granules that were sometimes used intervention was performed, per the authors. The patient to provide air contrast for an upper GI series. The authors presented to an emergency department (ED) the next day discovered that one BaSO4 product (E-Z-EM*) routinely with ongoing symptoms. The tablet was again demonstrated dissolved some of the ECASA tablets when prepared as a on radiographs, and again no intervention was performed full-strength solution. Half-strength E-Z-EM and a second before discharge. On the third day, the patient returned to commercial BaSO4 product (Malinckrodt) dissolved no the emergency room (ER) with evidence of acute on chronic tablets. The starting pH of the BaSO4 solutions did not seem salicylism, including a mixed acid–base disorder, tachypnea, to correlate with the ability to dissolve ECASA tablets. altered mental status, and an SSC of 110 mg/dL. A Hydrochloric acid and water solutions at pH ¼ 3, 5, and 7 gastrograffin-enhanced radiograph of the abdomen at this dissolved no tablets, whereas a NaHCO3 solution of pH ¼ 9 time showed no bezoar. The patient underwent hemodialysis rapidly dissolved all of the ECASA tablets. The addition of and survived. Subsequent endoscopy demonstrated chronic, effervescent granules (two different products tested) to the and likely preexisting, pyloric stenosis. The authors note BaSO4 solutions resulted in complete and rapid dissolution that BaSO4 was not known to degrade the CAP coating on of all ECASA tablets. Both of the effervescent granule ECASA tablets and postulated that repeated abdominal products used contained significant amounts of NaHCO3 and palpations may have caused the tablets to dissolve. were designed to release carbon dioxide when mixed with Halla et al.26 presented a case of a 29-year-old female water. Unfortunately, no pH measurements of the mixtures patient with a history of RA, vagotomy, and GOO, who containing both BaSO4 and effervescent granules are developed symptoms of acute salicylism with an initial SSC reported in the study. It is unclear whether the effervescent of 92 mg/dL. The patient was taking 8–12 Ecotrin tablets granules alkalinized the BaSO4 solutions or whether it was daily for over 9 months. She had complained of relapsing solely the effervescent action of the granules that enhanced abdominal pain, anorexia, early satiety, vomiting, and dissolution of the ECASA tablets. The authors made several weight loss. In addition to these symptoms, the patient also observations. First, there appears to be a syndrome of had relapsing episodes of confusion, forgetfulness, lethargy, anorexia, early satiety, and mild abdominal discomfort that and tinnitus. Shortly before hospital presentation, the patient many patients with ECASA bezoars report, even when signs was taking over-the-counter antacids and therapeutic doses of salicylate toxicity are absent. Second, Sogge et al.’s23 of Alka-SeltzerTM (322 mg aspirin per tablet, in addition to procedure appears safe and effective for the management of NaHCO3 and citric acid). On the day of her presentation to ECASA bezoars. And third, pH may not be the sole factor the hospital, the patient was lethargic, tachypneic, and that determines ECASA CAP-coating integrity. vomiting. The authors note that 21 intact Ecotrin tablets All of the salicylate bezoar reports involve some form of were recovered in the vomitus. No imaging or gastroscopy ECASA. All of the cases also feature chronic ingestion was performed. The patient underwent hemodialysis, and patterns in patients with a GOO. There were no reported

Clinical Toxicology vol. 49 no. 2 2011 Pharmacobezoars described and demystified 77 cases of ECASA bezoars forming after an acute overdose. Clomipramine Signs and symptoms of acute salicylism, as well as an Clomipramine is a chlorinated derivative of the prototypical elevated SSC, were found in some patients but not all. TCA imipramine. It is widely available in the US and abroad Vague GI complaints, such as early satiety and weight loss, for the treatment of obsessive compulsive disorder and are typical. It is important to keep in mind that ECASA depressive disorders. In the US, the drug is available in 25-, tablets remain intact for up to 11 days or potentially longer 50-, and 75-mg capsules as AnafranilTM or as a generic in the acidic environment of the stomach.22 ECASA tablets product. There are currently two case series describing a total are not radio-opaque in vivo, so oral contrast may be needed of six cases of pharmacobezoars involving AnafranilTM. 28 31 to effectively reveal gastric pill burdens. Because BaSO4 Lapostolle et al. described their experience with with effervescent granules may cause the rapid dissolution massive, acute AnafranilTM ingestions in four adult patients. of ECASA tablets in the stomach, gastrograffin may be a All four patients were attempting to commit suicide and had safer choice for diagnostic radiography and computed ingested multiple medications. Between 60 and 180 capsules tomography (CT). Endoscopy appears to be both a powerful of AnafranilTM were ingested in each case. All four patients diagnostic and treatment tool in this setting; however, underwent GL with a 40F lavage tube. All four patients had removing large pill burdens may be time consuming. GL plain radiographs of the chest demonstrating opaque masses appears to be unreliable, and whole bowel irrigation (WBI) interpreted as possible gastric pharmacobezoars. All four would not be expected to work well in a patient with a patients demonstrated features of TCA toxicity, and all had 23 GOO. The Sogge et al.’s procedure of NaHCO3 infusion/ laboratory evidence of clomipramine ingestion. Three aspiration via NG tube is a potential alternative to patients underwent endoscopy, and although pharmacobe- endoscopy for ECASA bezoar removal. There is insufficient zoars were identified in each case, all attempts at bezoar data to recommend Sogge et al.’s23 procedure over removal were unsuccessful. In addition, these three patients endoscopic bezoar removal at this time. developed significant upper GIH between 12 h and 7 days after endoscopy. The authors elected to avoid endoscopic Bromides manipulation of their fourth patient as a result of their Iberti et al.29 described a patient with prolonged sedation and experience with the other three, but the gastric opacity a steadily increasing serum bromide concentration (SBC) spontaneously resolved. The authors then demonstrated over the course 12 days despite GL and administration of a experimentally that 75-mg AnafranilTM capsules are radio- cathartic. A ‘‘gastric bezoar containing 0.3–0.5 cm identifi- opaque ex vivo. All four patients survived to discharge. able pill fragments’’ was discovered and removed at Lapostolle et al. concluded that clomipramine bezoars may endoscopy. A chemical analysis of the mass revealed an be detected with plain radiography and that endoscopic elevated bromide concentration. The authors included a manipulation of such pharmacobezoars may result in graph that demonstrated the sharp decrease in SBC that subsequent GIH. followed removal of the mass. The graph also demonstrated Hojer and Personne described two young adult patients that the patient’s SBC was stable until he or she was lavaged, who ingested 60 and 80 tablets, respectively, of 75-mg given charcoal, and given a sodium biphosphate–sodium slow-release clomipramine.32 In the first case, plain radio- phosphate mixture on day 5, after which the SBC rose graphy demonstrated a radio-opaque gastric mass. Copious sharply over the ensuing 7 days. It is unclear how or why the GL with a 30F lavage tube failed to clear the mass. An interventions on day 5 may have increased the rate of endoscope with a modified lavage tube taped to it was used bromide absorption in this patient. to fragment and clear the pharmacobezoar. Several passes of the instrument were required. In the second case, emergent Cholestyramine endoscopy was performed because physicians were unable Cohen et al.30 described the sole case of a cholestyramine to pass a lavage tube into the patient’s esophagus. bezoar. The patient was a 10-month-old female infant Endoscopy revealed a tenacious esophageal concretion of with congenital biliary atresia. She developed a proximal ‘‘tablet residues and granular material.’’ The obstructing large bowel obstruction after receiving 27 g of cholestyr- bezoar was removed by endoscopic fragmentation, flushing, amine over three consecutive days. BaSO4-enhanced and suction. In this case series, the ingestions were mixed, radiographs outlined an intraluminal mass at the site of and no confirmatory laboratory testing of the bezoars or the the obstruction. The child underwent laparotomy and was patient’s bodily fluids was performed. Neither patient found to have an abnormally rotated cecum and a large experienced seizures or hypotension nor did they develop impacted mass of a dry, granular substance. She expired 4 QRS widening on their electrocardiograms. Neither patient weeks later despite surgical decompression. The authors developed a GIH. The authors felt that the ingestion noted that the bezoar material did not dissolve in water, histories were reliable enough to implicate clomipramine acetylcysteine, or docusate sodium solutions and as the causative agent in both cases. They concluded that thus concluded that the substance was cholestyramine. early endoscopic bezoar removal contributed to the rapid It is difficult to determine the validity of the authors’ recovery and mild hospital courses of their patients. claim, as a fecal impaction may have presented in a The issue of whether or not clomipramine tablets are similar fashion. reliably radio-opaque in vivo is difficult to assess despite

Clinical Toxicology vol. 49 no. 2 2011 78 S.-E. Simpson these reports. There is evidence to suggest that clomipra- 7-h postingestion revealed 50–75 ferrous sulfate tablets, mine tablets are not reliably visible in small numbers on separate or in small clumps, throughout the stomach. plain radiographs.28 It is noteworthy that the bezoars Endoscopic tablet removal was deemed impractical and described in these cases appear to have been very dense imprudent given the multiple passes that would be required and difficult to remove, either by GL or endoscopic and diffuse mucosal injury already present. Gastrotomy was manipulation. Despite Lapostolle’s valid concerns regarding performed at 16 h to clear the remaining tablets, and the the safety of endoscopic management of tenacious bezoars, patient recovered. The authors termed the false radiographic leaving a large amount of such a toxic drug in place may be bezoar a ‘‘pseudoconcretion.’’ a more dangerous alternative. In a letter to the editor, Bock and Tenenbein37 suggested that WBI might obviate the need for surgical intervention in Iron patients like the one Foxford and Goldfrank encountered. In During the 1980s and 1990s, iron ingestions were a leading support of this position, Tenenbein reported a case series cause poisoning death in children. The tablets often had briefly describing the successful management of acute iron colorful sugar coatings that enticed toddlers. Iron salts, such ingestions in six patients (aged 2–19 years) with WBI.38 as ferrous sulfate, are not only widely available as Two patients received GL and ipecac, and two received only prescribed drugs, but they are also found in over-the- ipecac prior to WBI. Duration of WBI therapy ranged from counter nutritional supplements. A discussion of the toxicity 6 to 12 h, with all six patients passing ingested tablets in of various iron preparations is beyond the scope of this their stool. All of the patients tolerated the procedure and article, but the propensity for iron preparations, specifically recovered from their ingestions. Peak serum iron concentra- ferrous sulfate, to be retained in the stomach after overdose tions remained in the nontoxic range in the six patients. Two is well known. Retained gastric iron tablets are not clinically more recent case reports described prolonged WBI courses silent and may in fact enhance the toxicity of a given in toddlers and suggest that although the treatment tends to ingestion. Therefore, identification and removal of tablet be well tolerated, it can fail to completely clear tablets from burdens from the GI tract is a priority when managing acute the length of the GI tract.39,40 iron overdoses. Depending on the elemental iron content of Other successful approaches to managing perceived iron an iron preparation, plain radiographs can be used to detect tablet bezoars have been described. Tenenbein et al.41 retained tablets.28 Although many case reports exist that reported a case in which emergent gastrotomy was combined describe strategies for managing iron overdoses, only a with WBI at the intra- and postoperative periods in lieu of small number provide direct visualization of a patient’s GL. The authors stated that the child’s poor mental status, gastric contents to answer the question of whether or not a combined with her large gastric and enteric iron burden bezoar exists. encouraged the hybrid approach. Ng et al.42 described the Peterson and Fifield33 described a 15-month-old child endoscopic management of a true ferrous sulfate bezoar in an who presented 45 min after an acute ingestion of an adult patient who ingested 100 tablets 10 h prior to ED estimated 15–20 coated ferrous sulfate tablets, containing presentation. The bezoar was described as a 5-cm friable 60 mg of elemental iron per tablet. Initially, ipecac-induced mass that was mechanically fractured and removed in pieces. emesis following an oral dose of 5% sodium phosphate The majority of the cases described above do not describe solution delivered some tablet fragments. Abdominal radio- the presence of large iron tablet concretions. With the graphs revealed radio-opaque tablets in the stomach. exception of Ng et al.’s42 case, descriptions of retained Additional ipecac-induced emesis and GL with a 32F ferrous sulfate tablets typically describe granular slurries lavage tube failed to mobilize the tablets. Eight hours later, mixed with small tablet clumps. Depending on the elemental her serum iron concentration continued to rise, and the child iron content, iron tablets may or may not be radio-opaque. underwent emergency gastrotomy, whereupon a large Radiographs can therefore be misleading regarding the amount of gritty material was identified and scooped out presence of a bezoar. The well-documented ability of of the stomach. The authors stated that some of this material ferrous sulfate tablets to adhere to and injure GI mucosa, and remained imbedded in the gastric mucosa and was unable to thus enhance the systemic absorption of elemental iron, be removed despite vigorous irrigation. Nevertheless, the makes the distinction between an iron bezoar and retained child recovered. A discrete tablet concretion was not individual tablets clinically meaningless. In both cases, the described, but the authors speculated that the gelatin or iron burden can be difficult to remove. GL may remove sugar coating on ferrous sulfate tablets may have encour- some tablets, but additional interventions are often needed to aged bezoar formation. Venturelli et al.34 and Landsman accomplish satisfactory gastric decontamination. WBI, et al.35 described similar cases in 1982 and 1987, although apparently well tolerated, takes time to complete respectively. Foxford and Goldfrank36 described an adult and may not successfully mobilize larger bezoars or material who presented early after an acute massive ingestion of adherent to the GI tract mucosa. In addition, WBI may not ferrous sulfate and prenatal tablets containing iron. Radio- be a wise choice in patient who has evidence of graphy demonstrated a radio-opaque mass, suggesting a pylorospasm or deeply penetrating GI mucosal injury. The gastric bezoar. Spontaneous emesis and prolonged GL via use of endoscopy may be limited to those patients with 40F tube failed to completely clear the tablets. Endoscopy at manipulable bezoars. Because iron ingestions can cause

Clinical Toxicology vol. 49 no. 2 2011 Pharmacobezoars described and demystified 79 significant gastric mucosal injury, the risks and benefits of Meprobamate endoscopic manipulations must be carefully weighed. Meprobamate was a commonly prescribed minor tranquili- Gastrotomy, although invasive, seems to offer many benefits zer in the late 1950s. It was manufactured under several in the early management of acute iron ingestions. The trade names, including MiltownTM and EquanilTM, but also procedure allows for more thorough gastric decontamina- existed as generic form and in combination preparations. tion, excepting small adherent particles and intestinal iron The drug was available in 200-, 400-, and 600-mg strengths, burdens. Repeat abdominal radiographs and serum iron as both tablets and capsules. Meprobamate has since been concentration can be used to help gauge the effectiveness of superseded by drugs in the benzodiazepine class, as well as ongoing decontamination efforts (Figs. 1 and 2). its prodrug, carisprodol. As a result, it is very rare to encounter patients still taking meprobamate. Three case Magnesium oxide reports describing meprobamate bezoars were identified in Tatekawa et al.43 reported the case of a 26-year-old woman this review. with a history of chronic constipation and diarrhea who Jenis et al.44 reported the first case of a confirmed developed a SBO after taking moderate amounts of a meprobamate bezoar. The patient was a 51-year-old woman magnesium oxide cathartic over several months. Plain who took an estimated 38 g of the drug in a suicide attempt. radiographs demonstrated a radio-opaque mass in the pelvis. The patient was intubated for lethargy and underwent Subsequent laparotomy localized the mass to the ileum. multiple unsuccessful rounds of GL over 10 h via 30F Laboratory analysis of the 3 cm 6 3.5 cm mass revealed lavage tube. The patient’s initial blood meprobamate that it was composed primarily of magnesium oxide. The concentration was very elevated at 14.4 mg/dL. Within authors proposed that regular cathartic use had dehydrated 12 h of admission, the patient began to regain conscious- their patient and encouraged inspissation of the magnesium ness. She was extubated and transferred to the floor the same oxide. day, ambulatory and fully alert. The patient was subse-

Fig. 1. A plain radiograph of the abdomen of Velez et al.’s39 Fig. 2. A plain radiograph of Velez et al.’s39 patient obtained 36 h patient demonstrates radio-opaque ferrous sulfate tablets in the later, after the patient began to pass clear rectal efﬂuent during stomach and colon. The ﬁlm was obtained before WBI was started WBI. A persistent mass of tablets in the ascending colon is visible (permission granted by Marcel Dekker, Inc.). (permission granted by Marcel Dekker, Inc.).

Clinical Toxicology vol. 49 no. 2 2011 80 S.-E. Simpson quently found dead in her hospital bed 27 h after her initial option for patients who demonstrate life-threatening mepro- presentation. Autopsy revealed a 200 g mass with ‘‘the bamate toxicity, and in whom less invasive interventions consistency of semi-solid concrete’’ in the patient’s fail. The radio-opacity of the bezoar discovered by Davis stomach. Chemical analysis revealed that mass contained et al. is an interesting but isolated finding. The reliability of 25 g of meprobamate. Tissue concentrations of meproba- unenhanced radiography or CT in the detection of mate at the time of autopsy were 18 mg/dL in the blood, 14 meprobamate bezoars cannot be assessed at this time. To mg/dL in the brain, 30 mg/dL in the liver, 50 mg/dL in the date, there are no published reports of carisoprodol or kidneys, and 60 mg/dL in the fat. SomaTM bezoars in the MEDLINE database. The authors postulated that because meprobamate (a) is relatively insoluble in water, (b) is stable in acidic Nifedipine gastrointestinal therapeutic system environments, and (c) is capable of slowing gastric motility, The nifedipine gastrointestinal therapeutic system (GITS) is it is prone to forming bezoars in acute massive overdose. a controlled release nifedipine formulation that is in wide The authors stated that no additional pill containers were use today for the treatment of hypertension and stable found on the patient. Schwartz45 described the case of a angina. It is available under the trade names Procardia XL, 56-year-old woman who presented to the ED 4 h after an Adalat XL, and Nifedical XL in 30-, 60-, and 90-mg dosage acute ingestion of 36 g of meprobamate. The patient was forms. Procardia XL was approved by the Food and Drug intubated for worsening mental status and subsequently Administration (FDA) in 1989, and since then billions of underwent GL with 10 L of saline. She then received 30 g nifedipine GITS tablets have been manufactured and sold. of AC and a cathartic. The patient required 24 h of Obstructive nifedipine GITS bezoar formation, although vasopressor support and IV fluid resuscitation. A gastro- rare, is well described. The product insert for Procardia XL, scopy performed at 16 h demonstrated a large gelatinous along with those of other pharmaceutical products using mass in the stomach, which was left in place. Subsequent GITS technology, warns that patients with known GI hemodialysis decreased the patient’s blood meprobamate strictures should exercise caution when taking the pills. concentration and her level of sedation, but both rebounded The Alza Corporation (Mountain View, CA) developed within 4 h. A second gastroscopy demonstrated no change the GITS technology, under the trademark OROSTM in the size or location of the mass. At 40 h, the 140-g (osmotic-release oral systems), as a method to orally deliver gelatinous mass was removed by gastrotomy. The patient a relatively short acting drug at a constant rate over an again underwent hemodialysis and made a sustained extended time period. A nifedipine GITS tablet looks like an recovery. The mass was analyzed and found to contain ordinary, if not slightly large, drug tablet. It consists of a 24.9 g of meprobamate. A more recent case reported by nondeformable, water-permeable shell that surrounds an Davis et al.46 echoes features of Jenis et al.44’s case. Davis osmotically active bilayer core. The shell is made of et al.46 described a 45-year-old man being treated for self- cellulose acetate, and it is designed to remain intact despite inflicted stab wounds and a mixed meprobamate and the rigors of GI transit and the pressure exerted by the carbamazepine overdose. The patient was initially comatose swelling bilayer core. One side of the bilayer core is the and hypotensive but responded to supportive care measures. drug compartment. The other side contains polyethylene A CT angiogram of the chest incidentally revealed a radio- oxide, an inert, osmotically active polymer that expands opaque gastric mass, and subsequent endoscopy confirmed when exposed to water, and thus exerts pressure on the drug the presence of a gastric granular mass. The mass was compartment. A laser-drilled orifice in the shell permits the removed via endoscope, and chemical analysis revealed contents of the drug compartment to exit the tablet at a fixed both carbamezpine and meprobamate. The authors cited the rate. The rate of drug release is independent of surrounding patient’s low serum carbamazepine concentration as evi- pH or GI transit time. The depleted tablet shell is then dence that the bezoar primarily contained meprobamate. No excreted intact in the stool under normal circumstances. tissue meprobamate concentrations were reported. Finally, Shepherd48 reported a detailed case report of a 59-year- Eeckhout et al.47 briefly described the gastroscopic removal old man with a history of peptic ulcer disease who sought of ‘‘large [meprobamate] conglomerates’’ from a patient emergency care for relapsing abdominal pain and vomiting being treated for acute ingestion of 72 g of meprobamate. soon after starting nifedipine GITS of 90 mg. Endoscopy These case reports suggest that meprobamate can form revealed ‘‘a bezoar of tablets and tablet casings in the into a pharmacobezoar after massive ingestion. The stomach,’’ a large amount of tablet fragments in the insolubility of the drug combined with its ability to cause duodenum, as well as an active ulcer in the duodenal bulb. a paralytic ileus provides a plausible mechanism for bezoar The bezoar was not removed at that time. The patient was formation. The high-quality descriptions in the available discharged on a liquid diet and instructed to stop taking case reports combined with toxicodynamic clinical details nifedipine GITS. Two days later, the patient returned to the provided by Schwartz45 make a compelling case. The ED with ongoing abdominal pain and vomiting. A second physical descriptions of meprobamate bezoars suggest that endoscopy again demonstrated the bezoar. GL via Ewald GL and WBI may fail to clear such a concretion. tube was unsuccessful. The nifedipine GITS tablet concre- Endoscopic fragmentation and removal is a reasonable tion was endoscopically fragmented and removed, one tablet choice. Gastrotomy should be reserved as a decontamination at a time, in two separate sessions. Shepherd’s48 case is the

Clinical Toxicology vol. 49 no. 2 2011 Pharmacobezoars described and demystiﬁed 81

first that describes a nifedipine GITS bezoar and the practical difficulty of removing it. A similar case report followed in 1994, wherein a patient with a remote history of a Billroth II procedure developed weight loss, vomiting, and on endoscopic evaluation was shown to have a large mass of congealed nifedipine GITS tablets obstructing his gastric pouch.49 In this case also, the bezoar was fragmented and removed with multiple passes of an endoscope. In 1995, case reports of lower GI tract obstructive nifedipine GITS bezoars began to be reported. Georgopou- los and Gerdes described a patient with a history of colon cancer, hemicolectomy, and subsequent rectosigmoid stricture, who had a massive bezoar consisting of hundreds of 50 nifedipine GITS casings at the stricture site. A BaSO4 enema-enhanced radiograph, and a later a CT scan, suggested the diagnosis. Sigmoidoscopy confirmed a high- grade sigmoid stricture and a proximal bezoar. The stricture was dilated, and the patient successfully passed the concretion per rectum. This patient’s only symptoms were episodic abdominal pain and anorexia for 2 years. The patient exhibited no signs of a calcium channel antagonist toxidrome. Reid et al.51 described a mixed nifedipine GITS and sustained-release procainamide bezoar in a patient with a history of prior colectomy. A lower GIH complicated management of the bezoar. A combination of enemas, cathartics, and repeated sigmoidoscopic fragmentation was needed to relieve the obstruction. In this patient, the bezoar did not necessarily form at the site of a stricture but may have been precipitated by the anticholinergic effect of procainamide and the local smooth muscle relaxing effect of nifedipine. A final case by Smitz et al.52 described the delayed passage of over 60 nifedipine GITS shells in a patient with hypokalemia, atrial fibrillation, and ischemic colitis. A true obstruction caused by a bezoar, however, was not convincingly described. Small bowel nifedipine GITS bezoars also occur and present a unique challenge since colonscopy and endoscopy are not available as detection and management options. Kwon et al.53 reported the first such case. They described a patient with a massive (more than 300 shells) nifedipine GITS bezoar in a segment of ileum possibly narrowed by postsurgical adhesions. The concretion was detected by abdominal examination and oral contrast-enhanced radiography (Fig. 3). The patient eventually required a bowel Fig. 3. Abdominal radiographs of Kwon et al.’s53 patient resection to remove the bezoar but did well. In the same demonstrating a chronic nifedipine GITS bezoar in the jejunum. year, Taylor et al.54 reported a similar case as part of a broad Image A is an unenhanced film that reveals multiple-rounded topic review on medication-related bezoars. Their patient lucencies in the left upper and lower quadrants. Image B is of a had vague symptoms of weight loss, vomiting, and barium sulfate-enhanced radiograph, which shows characteristic tablet-shaped lucencies in the small bowel (permission granted by abdominal pain for several months before seeking emer- Springer Science and Business Media). gency care. A CT scan and ultrasound did not demonstrate a bezoar on the initial visit. The authors do not specify whether or not oral contrast was used. Nine months later, the patient returned to the ED with continued symptoms. A Wells et al.55 recently reported the case of a female patient second CT scan showed a segment of small bowel who intentionally overdosed on nifedipine GITS, extended thickening. Subsequent laparotomy revealed a small bowel release venlafaxine, atorvastatin, and sertraline. She was tumor and an associated obstructing nifedipine GITS bezoar. initially bradypneic and obtunded. Her initial heart rate and The affected bowel was resected and the patient recovered. blood glucose concentrations were not reported. The patient

Clinical Toxicology vol. 49 no. 2 2011 82 S.-E. Simpson was intubated, given 50 g of AC, IV NaHCO3, and supported with ‘‘vasopressors’’ for 36 h. The authors did not discuss this portion of the patient’s management in greater detail. She was thus stabilized and eventually transferred out of the intensive care unit, whereupon she quickly developed a calcium channel antagonist toxidrome (hypotension, bradycardia). Oral contrast-enhanced fluoro- scopy demonstrated the retained tablets in the stomach. NG lavage (tube size was not specified) failed to dislodge the tablets. The mass was eventually fragmented and removed by endoscopy (Figs. 4 and 5). The patient expired from hypotension and multisystem organ failure 3 days later. Nifedipine GITS obstructive bezoars have occurred in the esophagus, the stomach, the small bowel, and the large bowel. With the notable exception of Wells et al.’s55 case, the bezoars form under normal therapeutic conditions. The patients tend to be adults with a history of hypertension or angina. Some manner of GI tract dysmotility, usually caused by a stricture or physical narrowing, is a feature of nearly every case reported. However, in cases of massive acute overdose, or pediatric exploratory ingestion, nifedipine GITS can be retained in the stomach without preexisting Fig. 4. Oral contrast-enhanced CT image showing the large gastric GOO. With the exception of the esophageal obstruction collection of nifedipine GITS tablets in Wells et al.’s55 patient reports, the patients typically complain of vague GI (permission granted by the journal Pharmacotherapy). disturbances such as nausea, vomiting, abdominal pain, bloating, early satiety, anorexia, weight loss, diarrhea, and constipation. In some cases, the symptoms persist, and a bezoar is discovered long after the patient has stopped taking nifedipine GITS.56,57 Features of a calcium channel antagonist toxidrome are typically absent, and the tablets retrieved are largely depleted of nifedipine. GITS/OROSTM technology has been used or is currently in use in several other pharmaceutical products. AcutrimTM, ConcertaTM, Covera HSTM, Ditropan XLTM, DynaCirc CRTM, Efidac/24TM, Glucotrol XLTM, and OsmosinTM are or were at one time GITS formulations. Bass et al.58 reported a review article evaluating the GI safety of GITS/OROSTM products. The authors searched the available medical literature and used the Freedom of Information Act to obtain the FDA’s relevant postmarketing data on these products. Procardia XL was involved in more than 80% of the cases of GITS-induced obstruction. Clinical data regarding the cases reported to the FDA were not presented in the article. In the Fig. 5. The 167 nifedipine GITS tablets removed from the stomach TM medical literature; however, no other GITS/OROS of Wells et al.’s patient during endoscopy (permission granted by product has been shown to form bezoars. This discrepancy the journal Pharmacotherapy). is not completely explained by comparing tablet distribution data. For example, Bass et al.58 reported that 4.4 billion Procardia XL tablets were distributed between 1995 and 300 tablets of sustained-release potassium chloride. A radio- 1999, whereas 2.15 billion Glucotrol XL tablets distributed opaque mass was seen on plain radiography, which agrees in the same period. Yet the number of reported GI with experimental findings regarding the radio-opacity of obstructions attributed to Procardia XL was 19 times greater potassium chloride.28 Gastroscopy confirmed the presence than that of Glucotrol XL (95:5). None of the Glucotrol XL of a ‘‘sticky, irretrievable mass’’ in the patient’s severely cases have been reported in the medical literature. inflamed stomach. No other clinical details regarding the management of the acute overdose were provided. The Potassium chloride patient apparently survived the event but eventually Peeters and van der Werf59 briefly described the discovery developed a severe GOO apparently as a result of the of a pharmacobezoar in a patient who acutely ingested about irritant effects of the bezoar.

Clinical Toxicology vol. 49 no. 2 2011 Pharmacobezoars described and demystiﬁed 83

Quinidine were born at 25–27 weeks gestational age and determined Tecklenburg et al.60 described a dramatic case of a 16- that five had experienced complications related to SPS or month-old boy who developed severe quinidine toxicity CPS therapy. These five patients had formed palpable or after ingesting 22 extended-release QuinidexTM tablets. radio-opaque gastric masses after as few as one dose of resin Despite GL via NG tube, AC, prolonged WBI, and a variety via NG tube. Three of the patients ultimately died, but the of enhanced elimination interventions, the patient’s clinical authors state that the presence of the bezoars had not altered condition worsened commensurate with rising serum their clinical courses. The authors concluded that bezoars quinidine concentrations. The authors found the suggestion seemed to form only in those premature infants with a of a left upper quadrant mass on plain radiographs near day functional ileus. The total amount of SPS or CPS 3. Subsequent ultrasound of the abdomen revealed a 3 cm administered did not seem to correlate with the development gastric mass. The mass was identified and removed at of a bezoar. Koneru et al.65 described a 23-week gestation endoscopy. The child’s serum quinidine concentration soon neonate who developed an obstruction after receiving 9 g of declined, and he made a full recovery. Although the bezoar SPS via oral and rectal routes over the first 3 days of life. was never chemically analyzed to determine its content, a Clinical obstruction developed on day of life 13, and gastric aspirate obtained just prior to endoscopy contained a radio-opaque densities were discovered throughout the GI high concentration of quinidine. The authors cite the child’s tract with plain radiographs. A series of water-soluble profound hemodynamic instability, the large number of pills contrast enemas were administered to the child starting 40 ingested, the relative insolubility of quinidine, and the days after the identification of the radio-opaque masses and drug’s anticholinergic properties as potential contributors to succeeded in mobilizing the masses. No chemical or the formation of the pharmacobezoar. microscopic analysis of the expelled masses was presented, however, to confirm that these were SPS bezoars. Sodium polystyrene sulfonate Although the drugs are routinely used in adults, SPS and Sodium polystyrene sulfonate (SPS) is a cation-exchange CPS bezoars are reported almost exclusively in critically ill resin that is in wide used worldwide as an orally and rectally neonates. Infants with a functional ileus appear to be at a administered drug to treat hyperkalemia. Trade names of higher risk. Bowel obstruction due to the presence of these available SPS preparations include KayexylateTM, Kalex- bezoars is rare. Despite the poor outcomes in many of ateTM, and KionexTM. The drug is available as a liquid resin these patients, it is not clear whether SPS/CPC bezoars and functions by exchanging sodium ions for potassium ions increase morbidity and mortality. Patients often succumb to in the GI tract. Calcium polystyrene sulfonate (CPS) is a respiratory ailments or to dysrhythmias induced by closely related product that liberates calcium ions instead of hyperkalemia. SPS/CPS bezoars appear to be radio-opaque, sodium ions in exchange for potassium. which may aid early detection. Management options for Sherman et al.61 reported two cases of possible bezoar SPS/CPS bezoars are limited by the small size and fragile formation in critically ill, premature infants. The two infants health of the patients typically affected. The use of developed radio-opaque masses temporally associated with conservative measures, such as discontinuation of NG the administration of SPS via NG tube to treat hyperkalemia. SPS/CPS administration in favor of rectal administration, is Both infants died from causes apparently unrelated to the a rational approach. Serial plain radiography appears to be masses, but autopsies were not performed. The masses were a useful tool for following the transit or resolution of therefore not confirmed to be SPS bezoars. Menke et al.62 radio-opaque SPS/CPS bezoars. reported the case of a 36-week gestation neonate who developed a sepsis-like syndrome and hyperkalemia shortly Sucralfate after birth. The infant developed a radio-opaque gastric mass Sucralfate is an aluminum salt of sucrose octasulfate. It and abdominal distention after receiving a total of 16 g of was FDA approved for the treatment of duodenal ulcers in SPS in a 20% sorbitol solution over 2 days. The infant 1981. When sucralfate is in an acidic environment, the expired 2 days later, and a 4 cm 6 3cm6 3 cm firm lump aluminum ions dissociate from the sulfated sucrose moiety. was found in the patient’s stomach at autopsy. The authors Ionized sulfated sucrose then binds to proteins at the base reported that laboratory analysis confirmed SPS as the of an ulcer and forms an adherent exudate that protects the predominant ingredient of the mass. Metlay and Klionsky63 ulcer from further corrosion. Sucralfate is not absorbed described a critically ill premature infant who developed a systemically to a significant degree. In the US, sucralfate is radio-opaque gastric mass after receiving a total of 3 g of available by the trade name, CarafateTM (Sanofi-Aventis, SPS via NG tube over the course of several hours. The NJ), as either a tablet or a suspension. The CarafateTM infant expired on day of life 5, and autopsy revealed a 3-cm product insert warns that bezoar formation is a rare but diameter mass. Microscopic examination suggested that the reported side effect. mass was made up of SPS crystals, squamous debris, and Algozzine et al.66 reported a case of a possible sucralfate Candida albicans pseudohyphae. Ohlsson and Hosking64 bezoar. The authors described a 79-year-old man suffering reported a retrospective chart review of all the neonates who from severe Guillain–Barre´ paralysis and respiratory failure received SPS or CPS in their unit over the preceding 6 who developed GI bleeding during his hospitalization. The months. The authors identified eight patients, all of whom patient received enteral feedings, sucralfate 1 g four times a

Clinical Toxicology vol. 49 no. 2 2011 84 S.-E. Simpson day (QID), magnesia/alumina suspension, tincture of opium, described the formation of a gastric bezoar in a critically ill ferrous sulfate elixir, and Lactobacillus acidopholus tablets preterm neonate receiving only sucralfate for an apparent all via NG tube for 37 days. When a feeding tube could not GIH and breast milk feedings via NG tube.73 The infant be passed into the stomach, the patient underwent gastro- presented with abdominal distention, palpable mass, episo- scopy, whereupon a ‘‘hard, putty-like, flaky mass was found dic bradycardia, apnea, and hypoxia. Ultrasonography and in the esophagus and proximal stomach.’’ The mass was plain radiography were used to identify the mass. The gradually fragmented and dislodged over the course of 5 bezoar was eventually cleared by switching the infant to days with an endoscope-mounted water jet. The authors parenteral nutrition and performing hourly GL with saline. reported that the mass was analyzed and confirmed to be The infant was not intubated while receiving sucralfate composed primarily of sucralfate. However, the laboratory therapy, nor was he ever treated with antacids, although the that performed the analysis rebutted, stating that their authors point out that infants tend to have gastric pH in the analysis could not confirm the composition of the mass.67 4.0–4.8 range. The concurrent use of an aluminum-containing antacid in It is not clear how sucralfate works under normal Algozzine et al.’s66 patient confounded the analysis. Thus, therapeutic conditions, and thus, it is difficult to explain the validity of Algozzine et al.’s66 case report was rendered what caused the bezoars described above to form. It is questionable. possible that sucralfate congeals the abundant proteins Additional reports of esophageal bezoars in patients found in enteral feeding formulas. Anderson et al.68 note receiving oral sucralfate followed, however.68–70 The that their bezoar sample was 29% protein, similar to the patients in Algozzine et al.’s66 and subsequent case reports protein content of their patient’s tube feed product. Krupp were critically ill, adult inpatients. They were typically et al.70 performed a chemical analysis of their bezoar and started on sucralfate as an inpatient for an active GIH. They found a similar protein content. They discussed the relative also tended to have some kind of neurologic ailment that contributions of sucralfate, tube feed material, and Al(OH)3 may have induced esophageal dysmotility or gastroesopha- gel in their bezoar sample by examining the aluminum:sul- geal reflux (Guillain–Barre´ syndrome,66,68 diabetic neuro- fur ratio in their sample. However, given that sucralfate is pathy69). The patients were typically intubated and received known to release its aluminum in an acidic environment, enteral feeding via gastric tube. Antacid medications, either conclusions based on aluminum content of a bezoar are 67 oral Al(OH)3 or IV H2-antagonists, were featured as speculative. Quigley et al. cautioned readers to avoid coingestants. None of these patients had any preexisting drawing false conclusions from mineral analysis of bezoars. esophageal pathology. Antacid therapy has been commonly invoked as a risk factor Three interesting cases are worth examining separately for the formation of phytobezoars, and elevated gastric pH because they deviate from the pattern described above. Hart may similarly contribute to sucralfate bezoar formation by et al. described a case of esophageal bezoar attributed to altering the solubility of coingested dietary proteins. The sucralfate therapy in a 48-year-old woman with leukemia.71 proximal esophageal bezoar described by Shueke et al.72 is The patient was started on a sucralfate suspension 1 g QID, likely a simple coagulum of sucralfate over an ulcer and not IV ranitidine, and oral nystatin suspension to empirically unlike what one might find covering a peptic ulcer under treat odynophagia. The patient developed a distal esopha- normal therapeutic circumstances. geal obstruction 3 days later. A characteristic putty-like It seems prudent to avoid the combination of high-protein bezoar was identified and disintegrated on endoscopy. The tube feeds, acid reduction therapy, and sucralfate in any authors sent samples of the irrigant fluid to the same lab critically ill patient. Symptoms of esophageal obstruction or used by Algozzine et al.66 and claim that ‘‘the precipitate gastric fullness should be investigated with endoscopy, was composed of sucralfate’’ without additional explana- when possible. In infants, symptoms are likely to be subtle tion. Hart et al.’s patient was not intubated and did not have and suggestive of airway obstruction, respiratory disease, or a gastric tube in place. Details of her diet during initial sepsis. Ultrasonography and plain radiographs may be hospitalization were not provided. The patient was not noninvasive diagnostic options for infants. The tenacity and suffering from a neurologic ailment. Her esophagus high viscosity of these sucralfate-protein bezoars make them appeared normal at endoscopy, and she was able to tolerate difficult to dislodge. GL and WBI are not likely to succeed. ‘‘solid food’’ soon after the bezoar was dislodged. It is Endoscope-guided water jet dissolution may be a potentially unclear whether undiagnosed esophageal dysmotility, pre- safe and effective technique for managing softer cipitating dietary proteins, or medication interactions concretions. encouraged the formation of the bezoar in Hart et al.’s patient. A second unique case described by Shueke and Theophylline Mihas suggested that proximal esophageal obstruction can Theophylline is a methylxanthine drug that is in wide used also occur.72 The patient developed an obstructing bezoar at as an oral bronchodilator for patients with reactive airway the site of a preexisting proximal esophageal ulcer. Shueke diseases such as chronic obstructive pulmonary disease and and Mihas’ patient, like Hart et al.’s patient, was not asthma. In the US, the use of drug has diminished with the receiving enteral feeds at the time, nor was he intubated or introduction of alternatives, such as inhaled glucocorticoids suffering from a neurologic ailment. The third unique case and b2-agonists. However, theophylline is cheap to produce,

Clinical Toxicology vol. 49 no. 2 2011 Pharmacobezoars described and demystified 85 so it remains a popular drug in under-developed nations. Theophylline has been available worldwide in a large variety of immediate and sustained-release preparations. A small number of theophylline bezoar case reports exist in the English language medical literature. Cereda et al.74 described an 18-year-old female patient who ingested 60 Theo-durTM 300 mg tablets in an attempt to commit suicide. GL on arrival yielded no pill fragments. The patient then received supportive care including a dose of AC. Her initial serum theophylline concentration was 63 mg/mL. The patient’s clinical condition worsened over the next 12 h, and her serum theophylline concentration rose to 88 mg/mL. The patient underwent gastroscopy at that time, and a 2-cm diameter congealed mass of tablets was discovered and removed. The patient subsequently received oral mannitol and made a full recovery within another 12 h. Fig. 6. The waxy mass containing theophylline and charcoal Cereda et al.’s case report is brief but provides evidence that retrieved from the stomach of Bernstein et al.’s76 patient during removal of the tablet bezoar had significant and rapid impact autopsy (permission granted by Elsevier Limited). on the patient’s outcome. In response to Cereda et al.’s case report, a physician submitted a letter stating that upon reviewing theophylline toxicity cases in Cereda’s health district, there were at least four patients who apparently the diameter of the pills ingested or the inner diameter of the developed gastric theophylline bezoars (one patient vomited GL tube. The sustained-release theophylline tablets, in this the bezoar, two had bezoars detected at endoscopy, and one case, apparently had not coalesced into a waxy mass. at necropsy).75 The letter does not specify whether massive Sustained-release theophylline preparations, specifically ingestions occurred, nor does it specify which theophylline Theo-durTM tablets, have the potential to form gastric preparations were involved. bezoars in massive acute overdose. Even if the tablets do not Bernstein et al.76 reported a tragic but instructive case of form concretions, traditional means of gastric decontamina- fatal theophylline overdose complicated by gastric bezoar tion can fail, and the consequences can be severe. formation. In this case, a 54-year-old woman was seen and Theophylline is not radio-opaque,28 although oral contrast- treated in the ED for an acute theophylline overdose. GL via enhanced radiography or CT may suggest the presence of a Ewald tube with 10 L of normal saline yielded no pill bezoar. However, timely endoscopic examination of the fragments. The patient then received 50 m of AC and some stomach allows for both the positive identification and magnesium citrate before being discharged home without a immediate removal of retained drug. Prolonged observation period of observation. The patient subsequently expired at and consideration for additional GI tract decontamination home 8 h later. Autopsy revealed a large waxy mass in her with MDAC and WBI are prudent even after the successful stomach, calculated to contain approximately 29 g of removal of a gastric bezoar, as the detection of intestinal theophylline (Fig. 6). This case reinforces Cereda et al.’s74 tablets is technically difficult. findings that Theo-durTM can form gastric bezoars in acute overdose and that repeated attempts at GL can fail to clear Venlafaxine them. The tenacious consistency of Bernstein et al.’s76 Two cases describing probable venlafaxine pharmacobe- bezoar is visually documented in the original case report. zoars have been recently reported. The first by Djogovic The importance of prolonged close observation after et al.78,79 (described in both a case report and a letter to the sustained-release theophylline overdose is clearly demon- editor in 2007) occurred in a 47-year-old woman after a strated. massive, purposeful ingestion of up to 15 g of sustained- Saeki et al. reported a case of sustained-release theophyl- release venlafaxine (Effexor XRTM). The patient had line overdose in a 28-year-old man.77 A CT scan presented to the ED with new onset seizures and no demonstrated ‘‘many tablet shaped shadows in the stomach historical indication of an overdose. She was admitted to a and small intestine.’’ Although the authors do not specify general medical floor after being deemed clinically stable, whether any oral contrast had been used, the images they despite her continued drowsiness. The patient had another provided suggest that no oral contrast was used (Fig. 7). The seizure after admission, this time complicated by aspiration tablets appear to be radio-opaque in this case. Thirty-one and respiratory failure. After 5 days in intensive care, individual tablets were removed from the patient’s stomach endoscopy was performed and revealed a large gelatinous at endoscopy, which was performed within 130 min of ED mass of tablets. GL failed to remove the mass. It is unclear presentation. The patient made a full recovery with IV whether WBI was attempted or not. The mass of tablets was hydration and MDAC. The authors suggested that GL failed eventually removed via multiple passes with an endoscopy because of the size and number of tablets but did not report basket. The patient unfortunately remained in a persistent

Clinical Toxicology vol. 49 no. 2 2011 86 S.-E. Simpson

Fig. 7. Abdominal CT and endoscopic images of Saeki et al.’s patient demonstrate sustained-release theophylline tablets in the stomach and small bowel (permission granted Elsevier Limited).

vegetative state and her family withdrew care. In reading the reasonable and effective tool for the management of these two accounts of the same case, it is unclear whether the reported bezoars. When confronted with a critically ill tablets had formed a concretion or not. No chemical analysis venlafaxine overdose patient, urgent investigation for a was performed on the recovered material, and serial serum pharmacobezoar via noninvasive means is prudent. venlafaxine concentrations were not obtained during the hospitalization. In a second case report, Jones et al.80 briefly Verapamil described a 55-year-old woman who developed symptoms Rankin and Edwards81 reported a brief report of a fatal of GOO while being treated for a massive venlafaxine massive ingestion of sustained-release verapamil (Isoptin overdose. Because the case is presented as an abstract only, SR) in a 56-year-old man. The patient reportedly ingested there are very few clinical details. The authors stated that a the tablets 24–36 h prior to seeking care. Gastric deconta- CT scan confirmed their suspicion of a gastric pharmaco- mination was not performed. Despite pressor support with bezoar and that subsequent endoscopy succeeded in dopamine and isoprenaline, the patient expired 20 h after removing it. The authors reported a temporary improvement arrival with an asystolic cardiac arrest. At autopsy, the in the patient’s hemodynamic status after bezoar removal. patient had an elevated blood verapamil level and ‘‘matted An elevated serum venlafaxine concentration confirmed the lump of tablets, slightly larger than a golf ball’’ in his patient’s exposure to the drug. It is not clear whether the stomach. No laboratory analysis of the mass was reported. patient ingested a sustained-release preparation or not. Buckley et al.82 reported on their experience with three Unfortunately, this patient did not recover neurologic cases of slow-release verapamil overdose. The second case is function and her family withdrew care as well. included in this review because there is some evidence of There is currently enough information available to pharmacobezoar formation. The patient was a 23-year-old recommend that clinicians anticipate gastric pharmacobe- woman who purposefully ingested 4.8 g of slow-release zoar formation in patients who have acutely ingested large verapamil. She was asymptomatic on arrival to the ED and amounts of venlafaxine. It is unclear whether the venlafax- promptly underwent GL with unspecified results. WBI was ine bezoars are radio-opaque or not, therefore endoscopy then started, and after 2 h and 3.5 L of irrigation with would appear to be the diagnostic tool of choice. The failure polyethylene glycol, she passed ‘‘a conglomerate of tablets of GL to clear the venlafaxine bezoars in both cases suggests about 2-3 cm in diameter consistent with the tablets taken.’’ that the tablets may form adherent or tenacious concretions. The patient developed mild bradycardia and a mildly elevated In these two cases, it is unlikely that WBI would have been serum verapamil concentration but otherwise made a full any more efficient or reliable than GL. Endoscopy was a recovery. No chemical analysis of the bezoar was reported.

Clinical Toxicology vol. 49 no. 2 2011 Pharmacobezoars described and demystiﬁed 87

Sporer and Manning83 described a case of fatal acute Risk factors that predict a higher likelihood of verapamil ingestion in a 25-year-old man. The patient pharmacobezoar formation are particular to agent ingested. presented to the ER moderately hypotensive, with a normal Acuity and scale of the ingestion is of primary concern, heart rate, complaining of abdominal pain. Initial manage- because larger ingestions intuitively favor bezoar forma- ment included crystalloid blousing, AC administration, brief tion. Patient factors such as preexisting GOO, GI tract WBI, 2 g of calcium gluconate, dopamine, and epinephrine strictures, history of GI surgery, paralytic ileus, dehydra- infusions. The authors reported that mass spectroscopy of tion, and critical illness are also important to identify. the patient’s blood confirmed the presence of verapamil but Pharmaceutical agent factors, both intrinsic (tablet size, did not report a concentration. The patient steadily persistance, adhesiveness, viscosity) and extrinsic (antic- developed a worsening verapamil toxidrome over the next holinergic or opiate effects, ability to cause pylorospasm or 10 h, and eventually expired despite cardiac pacing and GI tract atony) should also be considered. There is no bolusing of glucagon. At autopsy, a ‘‘tennis ball-sized consistent overlap of risk factors among all the pharma- concretion of verapamil’’ was discovered in the patient’s cobezoars discussed here, however. Pharmacobezoars can stomach. The authors provided no other description or form both with sustained-release and immediate-release laboratory analysis of the bezoar. tablets, with liquid or nondigestible preparations, in normal or strictured GI tracts, and in massive overdose or in Combined: digoxin, metoprolol, and amitriptyline normal therapeutic settings. Robert et al.84 report a case of alleged massive acute Diagnostic options for the identification of pharmacobe- metoprolol, amitriptyline, and digoxin overdose in a 32- zoars should be tailored to individual patient and pharma- year-old woman who developed radiographic evidence of a ceutical agent factors. A thorough history and physical can gastric pharmacobezoar. The patient had a serum digoxin raise the suspicion of a pharmacobezoar. For example, concentration of 12.1 ng/mL approximately 4 h after chronic or relapsing vomiting, weight loss, and early satiety ingestion and was treated with digoxin antibodies. She in a patient taking nifedipine GITS or ECASA should never developed a TCA toxidrome or a beta-blocker prompt the consideration of an obstructing pharmacobezoar. toxidrome. The mass that was identified on plain radio- In the case of massive acute overdoses, worsening clinical graphs of the chest and abdomen was never confirmed by condition or rising serum drug concentrations after endoscopy. The patient received 50 g of charcoal every 4 h aggressive gastric decontamination should raise the possi- for 2 days and 4 L of polyethylene glycol WBI. The authors bility of a pharmacobezoar. However, toxicokinetic varia- concluded that the pharmacobezoar delayed the absorption tion among patients and toxicants limits the reliability of a of the ingested drugs and thus protected the patient. The ‘‘lab work’’ approach to diagnosing pharmacobezoars. authors’ interpretation of the radiographs is not adequately Moreover, experience with bromide29 and meprobamate44 supported by the available clinical evidence. indicates that falling serum drug concentrations or short- term improvements in clinical markers do not reliably exclude the presence of a pharmacobezoar after an acute Discussion overdose. Direct visualization of a pharmacobezoar allows Medications can accumulate in the GI tract for many reasons. for a measure of diagnostic certainty. Endoscopy and Therefore, establishing a clear definition of a pharmacobe- colonoscopy provide high-quality information but are often zoar is difficult. How does one distinguish between a limited to small areas of the GI tract. Conversely, pharmacobezoar and a handful of pills in the GI tract? Is radiography and CT, contrast enhanced or otherwise, can there a practical reason to make such a distinction? The sample much larger portions of the GI tract but can yield majority of pharmacobezoars discussed here appear to cause less reliable information. The choice of imaging also mechanical GI tract obstruction without systemic toxicity. depends on the location of the pharmacobezoar. Esopha- The predominance of obstructive pharmacobezoars in the geal, gastric and colonic bezoars are typically accessible medical literature likely occurs because such bezoars often with fiber optic devices. Bezoars in the jejunum and ileum, force surgical or endoscopic intervention. Prolonged or however, must be imaged by other means. Surgical delayed toxicity attributable to a pharmacobezoar is an exploration provides reliable information, but as a diag- exceedingly rare event or at least a rarely observed one. nostic tool, it is typically reserved for patients with clinical Nevertheless, the reports of meprobamate, theophylline, signs of a high-grade bowel obstruction. Ultrasound may clomipramine, venlafaxine, and ferrous sulfate bezoars provide some limited diagnostic information in premature reinforce the concept that persistent concretions of pharma- neonates who cannot tolerate endoscopy. The role for ceutical products can form after an overdose and that these ultrasound exploration in diagnosing gastric pharmacobe- tablet aggregates can persist and influence the course of care. zoars merits future study. Furthermore, individual pharmacobezoars can present and Management or removal of pharmacobezoars is also behave in strikingly different ways depending on their highly dependent on patient and pharmaceutical agent contents. For example, a stable gastric ECASA bezoar can factors. For example, benign SPS bezoars in critically ill suddenly dissolve and precipitate acute salicylism when the neonates should be conservatively managed, whereas drug pH in the stomach rises above a certain threshold. eluting theophylline bezoars or obstructing AC bezoars

Clinical Toxicology vol. 49 no. 2 2011 88 S.-E. Simpson demand more aggressive interventions such as endoscopic 5. Merriman T, Stokes K. Small bowel obstruction secondary to or surgical removal. In the majority of the cases reviewed, administration of activated charcoal. Aust N Z J Surg 1995; 65:288–289. 6. Gomez HF, Brent JA, Munoz DC 4th, Mimmack RF, Ritvo J, Phillips S, traditional means of gastric decontamination, such as GL, McKinney P. Charcoal stercolith with intestinal perforation in a patient WBI, ipecac-induced emesis, and cathartic administration, treated for amitriptyline ingestion. J Emerg Med 1994; 12:57–60. cannot be relied on to completely remove suspected or 7. Mizutani T, Naito H, Oohashi N. Rectal ulcer with massive confirmed pharmacobezoars. In addition, assessing the haemorrhage due to activated charcoal treatment in oral organopho- effectiveness of such decontamination procedures is sphate poisoning. Hum Exp Toxicol 1991; 10:385–386. 8. Goulbourne KB, Cisek JE. Small-bowel obstruction secondary to difficult. GL aspirations and WBI effluent that initially activated charcoal and adhesions. Ann Emerg Med 1994; 24:108–110. yield ingested material but then turn clear can mislead one 9. Mauro LS, Nawarskas JJ, Mauro VF. Misadventures with activated into believing that gastric decontamination was effective in charcoal and recommendations for safe use. Ann Pharmacother 1994; clearing the GI tract. Jenis et al.’s44 meprobamate case, 28:915–924. among others, highlights the dangers of misinterpreting 10. Flores F, Battle WS. Intestinal obstruction secondary to activated charcoal. Contemp Surg 1987; 30:57–59. normal gastric aspirates. When the potential for delayed or 11. Chan JC, Saranasuriya C, Waxman BP. Bezoar causing small bowel recurrent toxicity from a pharmacobezoar exists, prolonged obstruction after repeated activated charcoal administration. Med J observation of a patient in a high-acuity setting is prudent. Aust 2005; 183:537. 12. Dorrington CL, Johnson DW, Brant R; for the MDAC Study Group. The frequency of complications associated with the use of multiple- Conclusion dose activated charcoal. Ann Emerg Med 2003; 41:370–377. 13. Position statement and practice guidelines on the use of multi-dose Pharmacobezoars enjoy a near mythological status. They can activated charcoal in the treatment of acute poisoning. American form under a wide variety of circumstances and in a wide Academy of Clinical Toxicology; European Association of Poisons variety of patients. Yet they are rare entities that are often Centres and Clinical Toxicologists. J Toxicol Clin Toxicol 1999; difficult to diagnose. The clinical impact of a given 37:731–751. 14. Havens W. Intestinal obstruction caused by colloidal aluminum pharmacobezoar is also quite variable and sometimes difficult hydroxide. JAMA 1939; 113:1564–1565. to assess. Rules for suspecting, diagnosing, and properly 15. Townsend CM Jr, Remmers AR Jr, Sarles HE, Fish JC. Intestinal managing a pharmacobezoar are highly dependent on the obstruction from medication bezoar in patients with renal failure. N pharmaceutical agent or agents involved. Becoming familiar Engl J Med 1973; 288:1058–1059. with the sparse data available on pharmacobezoars and 16. Korenman MD, Stubbs MB, Fish JC. Intestinal obstruction from medication bezoars. JAMA 1978; 240:54–55. maintaining a high index of suspicion in future clinical 17. Girotti MJ, Ruddan J, Cohanim M. Amphojeloma: antacid impaction encounters may be the best way to improve diagnostic in a critically ill patient. Can J Surg 1984; 27:379–380, 382. sensitivity and accuracy. Moving forward, there are many 18. Potyk D. Intestinal obstruction from impacted antacid tablets. N Engl avenues of study that may increase our understanding about J Med 1970; 283:134–135. how pharmacobezoars form in the acute overdose setting. The 19. Burruss GL, Van Voorst SJ, Crawford AJ, et al. Small bowel obstruction from an antacid bezoar: a ranitidine-antacid interaction? role that particular pharmaceutical excipients play in the South Med J 1986; 79:917–918. formation of bezoars is largely unexplored. Similarly, the role 20. Portuguez-Malavasi A, Aranda JV. Antacid bezoar in a newborn. of radiographic, endoscopic, or ultrasonographic screening of Pediatrics 1979; 63:679–680. high-risk patients requires greater study. As our shared 21. Baum J. Enteric coated aspirin today. J Rheumatol 1985; 12:829–832. experience with individual pharmaceutical products increases 22. Harris FC. Pyloric stenosis: hold-up of enteric coated aspirin tablets. Br J Surg 1973; 60:979–981. over time, we may better understand their propensity to form 23. Sogge MR, Griffith JL, Sinar DR, Mayes GR. Lavage to remove bezoars and the conditions required for such a transformation. enteric-coated aspirin and gastric outlet obstruction. Ann Intern Med 1977; 87:721–722. 24. Sherry E. Pyloric obstruction caused by aspirin: case report. N Z Med Declaration of interest J 1979; 89:174–175. 25. Springer DJ, Groll A. Poisoning with enteric-coated acetylsalicylic The authors report no conflicts of interest. The authors alone acid complicating gastric outlet obstruction. Can Med Assoc J 1980; are responsible for the content and writing of the paper. 122:1032–1034. 26. Halla JT, Fallahi S, Hardin JG. Acute and chronic salicylate intoxication in a patient with gastric outlet obstruction. Arthritis References Rheum 1981; 24:1205–1207. 27. Bogacz K, Caldron P. Enteric-coated aspirin bezoar: elevation of 1. Andrus CH, Ponsky JL. Bezoars: classification, pathophysiology, and serum salicylate level by barium study. Case report and review of treatment. Am J Gastroenterol 1988; 83:476–478. medical management. Am J Med 1987; 83:783–786. 2. Watson WA, Cremer KF, Chapman JA. Gastrointestinal obstruction 28. Savitt DL, Hawkins HH, Roberts JR. The radiopacity of ingested associated with multiple-dose activated charcoal. J Emerg Med 1986; medications. Ann Emerg Med 1987; 16:331–339. 4:401–407. 29. Iberti TJ, Patterson BK, Fisher CJ Jr. Prolonged bromide intoxication 3. Ray MJ, Radin DR, Condie JD, Halls JM. Charcoal bezoar. Small- resulting from a gastric bezoar. Arch Intern Med 1984; 144:402–403. bowel obstruction secondary to amitriptyline overdose therapy. Dig 30. Cohen MI, Winslow PR, Boley SJ. Intestinal obstruction associated Dis Sci 1988; 33:106–107. with cholestyramine therapy. N Engl J Med 1969; 280:1285–1286. 4. Atkinson SW, Young Y, Trotter GA. Treatment with activated 31. Lapostolle F, Finot MA, Adnet F, Borron SW, Baud FJ, Bismuth C. charcoal complicated by gastrointestinal obstruction requiring sur- Radiopacity of clomipramine conglomerations and unsuccessful gery. BMJ 1992; 305:563. endoscopy: report of 4 cases. J Toxicol Clin Toxicol 2000; 38:477–482.

Clinical Toxicology vol. 49 no. 2 2011 Pharmacobezoars described and demystiﬁed 89

32. Hojer J, Personne M. Endoscopic removal of slow release 59. Peeters JW, van der Werf SD. Gastric stenosis after potassium clomipramine bezoars in two cases of acute poisoning. Clin Toxicol chloride ingestion. Endoscopy 1998; 30:S110. (Phila) 2008; 46:317–319. 60. Tecklenburg FW, Thomas NJ, Webb SA, Case C, Habib DM. 33. Peterson CD, Fifield GC. Emergency gastrotomy for acute iron Pediatric ECMO for severe quinidine cardiotoxicity. Pediatr Emerg poisoning. Ann Emerg Med 1980; 9:262–264. Care 1997; 13:111–113. 34. Venturelli J, Kwee Y, Morris N, Cameron G. Gastrotomy in the 61. Sherman S, Friedman AP, Berdon WE, Haller JO. Kayexalate: a management of acute iron poisoning. J Pediatr 1982; 100:768–769. new cause of neonatal bowel opacification. Radiology 1981; 138:63– 35. Landsman I, Bricker JT, Reid BS, Bloss RS. Emergency gastrotomy: 64. treatment of choice for iron bezoar. J Pediatr Surg 1987; 22:184–185. 62. Menke JA, Stallworth RE, Binstadt DH, Strano AJ, Wallace SE. 36. Foxford R, Goldfrank L. Gastrotomy – a surgical approach to iron Medication bezoar in a neonate. Am J Dis Child 1982; 136:72–73. overdose. Ann Emerg Med 1985; 14:1223–1226. 63. Metlay LA, Klionsky BA. An unusual gastric bezoar in a 37. Bock GW, Tenenbein M. Whole bowel irrigation for iron overdose. newborn: polystyrene resin and Candida albicans. J Pediatr 1983; Ann Emerg Med 1987; 16:137–138. 102:121–123. 38. Tenenbein M. Whole bowel irrigation in iron poisoning. J Pediatr 64. Ohlsson A, Hosking M. Complications following oral administration 1987; 111:142–145. of exchange resins in extremely low-birth-weight infants. Eur J 39. Velez LI, Gracia R, Mills LD, Shepherd G, Feng SY. Iron bezoar Pediatr 1987; 146:571–574. retained in colon despite 3 days of whole bowel irrigation. J Toxicol 65. Koneru P, Kaufman RA, Talati AJ, Jenkins MB, Korones SB. Clin Toxicol 2004; 42:653–656. Successful treatment of sodium polystyrene sulfonate bezoars with 40. Kaczorowski JM, Wax PM. Five days of whole-bowel irrigation serial water-soluble contrast enemas. J Perinatol 2003; 23:431–433. in a case of pediatric iron ingestion. Ann Emerg Med 1996; 27:258–263. 66. Algozzine GJ, Hill G, Scoggins WG, Marr MA. Sucralfate bezoar. N 41. Tenenbein M, Wiseman N, Yatscoff RM. Gastrotomy and whole bowel Engl J Med 1983; 309:1387. irrigation in iron poisoning. Pediatr Emerg Care 1991; 7:286–288. 67. Quigley MA, Flicker M, Caldwell EG. Sucralfate bezoar – theory or 42. Ng HW, Tse ML, Lau FL, Chu W. Endoscopic removal of iron bezoar fact? Am J Gastroenterol 1986; 81:724–725. following acute overdose. Clin Toxicol (Phila) 2008; 46:913–915. 68. Anderson W, Weatherstone G, Veal C. Esophageal medication bezoar 43. Tatekawa Y, Nakatani K, Ishii H, Paku S, Kasamatsu M, Sekiya N, in a patient receiving enteral feedings and sucralfate. Am J Nakano H. Small bowel obstruction caused by a medication bezoar: Gastroenterol 1989; 84:205–206. report of a case. Surg Today 1996; 26:68–70. 69. Carrougher JG, Barrilleaux CN. Esophageal bezoars: the sucralith. 44. Jenis EH, Payne RJ, Goldbaum LR. Acute meprobamate poisoning. A Crit Care Med 1991; 19:837–839. fatal case following a lucid interval. JAMA 1969; 207:361–362. 70. Krupp KB, Johns P, Troncoso V. Esophageal bezoar formation in a 45. Schwartz HS. Acute meprobamate poisoning with gastrotomy and tube-fed patient receiving sucralfate and antacid therapy: a case report. removal of a drug-containing mass. N Engl J Med 1976; 295:1177–1178. Gastroenterol Nurs 1995; 18:46–48. 46. Davis RN, Rettmann JA, Christensen B. Relapsing altered mental status 71. Hart RS, Levin B, Gholson CF. Esophageal obstruction caused by secondary to a meprobamate bezoar. J Trauma 2006; 61:990–991. sucralfate impaction. Gastrointest Endosc 1989; 35:474–475. 47. Eeckhout E, Huyghens L, Loef B, Maes V, Sennesael J. Meprobamate 72. Shueke M, Mihas AA. Esophageal bezoar due to sucralfate. poisoning, hypotension and the Swan-Ganz catheter. Intensive Care Endoscopy 1991; 23:305–306. Med 1988; 14:437–438. 73. Strozik KS, Walele AH, Hoffman H. Bezoar in a preterm baby 48. Shepherd MF. Bezoar formed by fragments of extended-release associated with sucralfate. Clin Pediatr (Phila) 1996; 35:423–424. nifedipine tablets. Clin Pharm 1993; 12:849–852. 74. Cereda JM, Scott J, Quigley EM. Endoscopic removal of pharmaco- 49. Stack PE, Patel NR, Young MF, Ferslew KE, Thomas E. bezoar of slow release theophylline. Br Med J (Clin Res Ed) 1986; Pharmacobezoars – the irony of the antidote: first case report of 293:1143. nifedipine XL bezoar. J Clin Gastroenterol 1994; 19:264–265. 75. Smith WDF. Endoscopic removal of a pharmacobezoar of slow 50. Georgopoulos S, Gerdes H. Retention of nifedipine extended release release theophylline. Br Med J (Clin Res Ed) 1987; 294:125. tabs in a patient with a colonic stricture. Am J Gastroenterol 1995; 76. Bernstein G, Jehle D, Bernaski E, et al. Failure of gastric emptying 90:2224–2226. and charcoal administration in fatal sustained-release theophylline 51. Reid T, Rubins JB, Levine J, DeCelles J, Silvis S. Colonic medication overdose: pharmacobezoar formation. Ann Emerg Med 1992; bezoar from extended-release nifedipine and procainamide. Arch Fam 21:1388–1390. Med 1995; 4:715–717. 77. Saeki S, Shimoda T, Sakai H, Soejima Y, Matsuse H, Kohno S. 52. Smitz S, Bonnet V, Delporte JP. Severe gastrointestinal dysfunction Successful treatment of theophylline toxicity by upper gastrointestinal and retention of extended release nifedipine tablets. J Am Geriatr Soc endoscopy. Respir Med 2003; 97:734–735. 1998; 46:656–657. 78. Djogovic D, Hudson D, Jacka M. Gastric bezoar following 53. Kwon HY, Scott RL, Mulloy JP. Small bowel Procardia XL tablet venlafaxine overdose. Clin Toxicol (Phila) 2007; 45:735. bezoar mimicking cystic pneumatosis intestinalis. Abdom Imaging 79. Hudson D, Djogovic D. Alternate venlafaxine kinetics in overdose 1996; 21:142–144. patients. J Clin Pharm Ther 2007; 32:109. 54. Taylor JR, Streetman DS, Castle SS. Medication bezoars: a literature 80. Jones JB, Moore JC, Setzer SC, Lintner CP, Roberts DJ. Massive review and report of a case. Ann Pharmacother 1998; 32:940–946. venlafaxine ingestion resulting hypotension and pharmacobezoar 55. Wells CD, Luckritz TC, Rady MY, Zornik JM, Leighton JA, Patel formation, in 2009 North American Congress of Clinical Toxicology BM. Bezoar formation requiring endoscopic removal after intentional (NACCT). Clin Toxicol (Phila) 2009; 47:747. Abstracts. overdose of extended-release nifedipine. Pharmacotherapy 2006; 81. Rankin RJ, Edwards IR. Overdose of sustained release verapamil. N Z 26:1802–1805. Med J 1990; 103:165. 56. Niezabitowski LM, Nguyen BN, Gums JG. Extended-release 82. Buckley N, Dawson AH, Howarth D, Whyte IM. Slow-release nifedipine bezoar identified one year after discontinuation. Ann verapamil poisoning. Use of polyethylene glycol whole-bowel lavage Pharmacother 2000; 34:862–864. and high-dose calcium. Med J Aust 1993; 158:202–204. 57. Yeen WC, Willis IH. Retention of extended release nifedipine 83. Sporer KA, Manning JJ. Massive ingestion of sustained-release capsules in a patient with enteric stricture causing recurrent small verapamil with a concretion and bowel infarction. Ann Emerg Med bowel obstruction. South Med J 2005; 98:839–842. 1993; 22:603–605. 58. Bass DM, Prevo M, Waxman DS. Gastrointestinal safety of an 84. Robert R, Frat JP, Veinstein A, Rouffineau J. Protective effect of extended-release, nondeformable, oral dosage form (OROS): a medication bezoar after a massive beta-blocker, digoxin, and retrospective study. Drug Saf 2002; 25:1021–1033. amitriptyline poisoning. Clin Toxicol (Phila) 2005; 43:381–382.

Clinical Toxicology vol. 49 no. 2 2011