Cambridge Protech, USA). The analysis was performed atthe Department of Microbiology and Immunology at ourhospital. Enteric pathogens were not demonstrable. Bloodcultures were negative. X-rays of chest and abdomen werenormal. Colonoscopy was not performed.

Treatment with systemic metronidazole was given for 5days, with rapid improvement of the patient’s condition.However, 2 wk later she was readmitted with the samesymptoms, andC. difficile toxin A was again demonstratedin feces. Treatment with metronidazole for 10 days resultedin complete remission of all symptoms and signs.

DISCUSSION

Clindamycin, cephalosporins, and penicillins are the mostfrequently reported causes of AAD. However, nearly anyantibiotic may induce this condition (1). Gerdinget al. (1)have suggested three main diagnostic criteria for enterotoxicdiarrhea: 1) Frequency of watery stools.6 in 36 h; 2)presence ofC. difficile toxin in stool, or endoscopicallyverified pseudomembranes; and 3) exclusion of other causesof diarrhea. Our patient developed watery diarrhea with afrequency of about 20 stools/day until admission.C. difficiletoxin was demonstrated in feces and no other cause ofdiarrhea was identified. Blood was present in feces, sug-gesting not onlyC. difficile-associated diarrhea, but also acolitis. However, colonoscopy was not performed. In acase-control study by Laiet al. (5), only exposure to anti-biotics and prior respiratory tract infection were found to besignificant risk factors forC. difficile-associated diarrhea.The only antibiotic used by our patient within the last 3months before the start of diarrhea was vaginal applicationof clindamycin. She had no sign of respiratory tract infec-tion.

The bioavailability after use of 2% clindamycin vaginalcream has been shown to be around 3% (2.7–4.7%), bothfor symptomatic patients and for healthy controls (6). Thisindicates that systemic exposure after intravaginal treatmentwith clindamycin is low. Accordingly, long lasting localtreatment has been considered to be necessary to developC.difficile-associated diarrhea (4, 7). However, we concludethat even short term application of clindamycin vaginalcream may result in AAD.

Reprint requests and correspondence:K. Vikenes, M.D., De-partment of Medicine, Haukeland Hospital, N-5021, Norway.

Received Sep. 15, 1998; accepted Jan. 11, 1999.

REFERENCES

1. Gerding DN, Johnson S, Peterson LR, et al. Clostridium diffi-cile-associated diarrhea and colitis. Infect Control Hosp Epide-miol 1995;16:459–77.

2. Milstone EB, McDonald AJ, Scholhamer CF. Pseudomembra-nous colitis after topical application of clindamycin. Arch Der-matol 1981;117:154–5.

3. Parry MF, Rha CK. Pseudomembranous colitis caused by top-ical clindamycin phosphate. Arch Dermatol 1986;122:583–4.

4. Trexler MF, Fraser TG, Jones MP. Fulminant pseudomembra-nous colitis caused by clindamycin phosphate vaginal cream.Am J Gastroenterol 1997;92:2112–3.

5. Lai KK, Melvin ZS, Kotilainen HR, et al. Clostridium difficileassociated diarrhea: Epidemiology, risk factors, and infectioncontrol. Infect Control Hosp Epidemiol 1997;18:628–32.

6. Borin MT, Powley GW, Tackwell KR, et al. Absorption ofclindamycin after intravaginal application of clindamycin phos-phate 2% cream. J Antimicrob Chemother 1995;35:833–41.

7. Siegle RJ, Fekety R, Sarbone PD et al. Effects of topicalclindamycin on intestinal microflora in patients with acne. J AmAcad Dermatol 1986;15:180–5.

Rapunzel Syndrome—A Case ReportSham L. Singla, M.S., F.I.C.S., K. N. Rattan, M.S., M.Ch.,Narender Kaushik, M.B.B.S., and S. K. Pandit, M.S., M.Ch.Departments of Surgery and Paediatric Surgery, Pt.B.D.SharmaPostgraduate Institute of Medical Sciences, Rohtak, India

ABSTRACTThe gastric trichobezoars usually occur in young girls, oftenthose with psychiatric disorders. Rarely these are known toextend from the stomach to the small intestine as a tail, whenthey are termed the Rapunzel syndrome. Until 1997, only 10such cases have been reported in the literature. We reportanother case in which we could extract the trichobezoar bygastrotomy and enterotomy. (Am J Gastroenterol 1999;94:1970–1971. © 1999 by Am. Coll. of Gastroenterology)

INTRODUCTION

The term “bezoar” has been derived from the wordbazahr(Arabic), meaning an antidote believed to have medicinalvalue (1). It is termed “trichobezoar” when a conglomera-tion is made up of swallowed hair. Rarely, these tricho-bezoars may extend from the stomach to the small intestineas a tail, when they have been named the Rapunzel syn-drome. Up to 1997, only 10 cases of this entity had beenreported (1–5). We wish to add another case to the literature,the second from India.

CASE REPORT

A 9-yr-old girl, weighing 22 kg was admitted with abdom-inal pain, vomiting, loss of appetite, and lump abdomen. Sheused to swallow hair, threads, and rags. On palpation a firm,slightly mobile, transversely placed intra-abdominal masswas present in the epigastrium. USG revealed an echogenicmass in the stomach with linear postacoustic shadowing.Diagnosis of a trichobezoar was made. Laparotomy revealeda large trichobezoar in the stomach, the tail of which wasextending up to the ileum (Fig. 1). It was removed byperforming a gastrotomy. While removing the tail, a jejunalperforation occurred, which was closed. The patient was

1970 Brief Case Reports AJG – Vol. 94, No. 7, 1999

discharged in healthy condition, to continue follow-up withthe psychiatrist.

DISCUSSION

Trichobezoars are commonly seen in young girls with psy-chiatric disorders. The hair swallowed collect in the stomachto form a hair ball, the interstices of which get filled by avariety of vegetable materials. The stomach is not able topush these hairs out of its lumen because the friction surfaceis insufficient for propulsion by peristalsis (6). Rarely, how-ever, these bezoars can get broken and can pass into theintestine to cause intestinal obstruction, ulceration, bleed-ing, and perforation. Small intestinal bezoars have also beenreported after truncal vagotomy and compression of theduodenum by the superior mesenteric artery (7, 8).

In the Rapunzel syndrome, Duncanet al. recommendedbezoar extraction by multiple enterotomies (2). Baliket al.could extract the jejunal extension of trichobezoar by string-ing the intestine along the length of the bezoar (3). Uroz andcoauthors extracted the jejunal extension via the gastrotomyitself (4). We also tried extraction as recommended by Uroz(4). During the maneuver, an indentation in the jejunum wasnoted at one site, and on further manipulation a perforationoccurred. Possibly the bezoar was adherent at this site. Toavoid such an iatrogenic perforation, we feel that, on ob-serving any indentation on the intestinal wall, one shouldimmediately resort to multiple enterotomies, as recom-mended also by Duncanet al. (2).

Reprint requests and correspondence:Dr. Sham L. Singla, 321,Housing Board Colony, Rohtak-124001, Haryana, India.

Received Sep. 11, 1998; accepted Dec. 11, 1998.

REFERENCES

1. Balik E, Ulman I, Taneli C, et al. The Rapunzel syndrome: Acase report and review of the literature. Eur J Pediatr Surg1993;3:171–3.

2. Duncan ND, Aitken R, Venugopal S, et al. The Rapunzelsyndrome. Report of a case and review of the literature. WestIndian Med J 1994;43:63–5.

3. Uroz Tristan J, Garcia Urgelles X, Melian Perez-Marin S.Rapunzel syndrome: A report of a new case. Cir Pediatr 1996;9:40–1.

4. Pul N, Pul M. The Rapunzel syndrome (trichobezoar) causinggastric perforation in a child: A case report. Eur J Pediatr1996;155:18–9.

5. Senapati MK, Subramanian S. Rapunzel syndrome. Trop Doct1997;27:53–4.

6. Santiago Sanchez CA, Garau Diaz P, Lugo Vicente HL. Tri-chobezoar in a 11-year old girl: A case report. Bol Asoc MedPR 1996;88:8–11.

7. Escamilla C, Robles-Campos R, Parrilla-Paricio P, et al.Intestinal obstruction and bezoars. J Am Coll Surg 1994;179:285– 8.

8. Doski JJ, Priebe CJ Jr, Smith T, et al. Duodenal trichobezoarcaused by compression of the superior mesentric artery. J Pe-diatr Surg 1995;30:1598–9.

Pseudomembranous ColitisAfter Itraconazole TherapyAdrienne J. Nguyen, M.D., Douglas B. Nelson, M.D., andJoseph R. Thurn, M.D.Department of Medicine VA Medical Center and University ofMinnesota, Minneapolis, Minnesota

ABSTRACTA 53-yr-old man was admitted with new onset of abdominalpain and nonbloody diarrhea 1 month after exposure to theantifungal agent itraconazole. Flexible sigmoidoscopy dem-onstrated the presence of pseudomembranes, and subse-quent evaluation excluded other causes of diarrhea. Disrup-tion of the resident fungal flora of the colon by itraconazoleis proposed as the mechanism by which this patient devel-oped pseudomembranous colitis. This association has notpreviously been reported. (Am J Gastroenterol 1999;94:1971–1973. © 1999 by Am. Coll. of Gastroenterology)

INTRODUCTION

Pseudomembranous colitis (PMC) was considered relativelyrare in the preantibiotic era when it was first described byFinney in 1893 (1). With the increasingly prevalent use ofantibiotics, particularly broad-spectrum antibiotics, PMChas become a major health problem. Nearly all antimicrobialagents with antibacterial activity have been implicated (2).We report the first case of pseudomembranous colitis asso-ciated with the antifungal agent itraconazole.

Figure 1. Figure showing the extracted specimen of trichobezoar.

1971AJG – July, 1999 Brief Case Reports