Biomedical subjects
S C Abraham
Publications and source records attributed to S C Abraham.
Florid decidual reaction mimicking gastrointestinal malignancy in a primipara woman.
Explore the source record for details and available documents.
The clinical spectrum of duodenal polyps in pediatrics.
OBJECTIVES: The aim of this retrospective study was to determine the prevalence, clinical presentation, and histologic subclassification of duodenal polyps identified on endoscopy (EGD) in pediatric patients. METHODS: We performed an 18-year retrospective study of all pediatric patients (< 21 years) with duodenal polyps diagnosed between 1983 and 2001 at The Johns Hopkins Children's Center. Our analysis includes a formal histologic evaluation of duodenal polyps either biopsied using cold-forceps or removed by snare cautery. RESULTS: Duodenal polyps were reported in 22 of 5766 EGDs (0.4%) performed in 16 (M:F; 1:1) patients with a mean (SD) age of 14.1 (5.1) years. Polyps were equal in both the Caucasian and African American population (adjusted ratio 1.2:1). The histologic subtypes included Adenomatous (42%), Brunner's gland hyperplastic (33%), hamartomatous (17%), and heterotopic gastric gland polyps (8%). The most frequent indication for EGD was surveillance in patients with polyposis syndromes; most of these patients were asymptomatic at the time of their EGD. In comparison, the most frequent indication for an EGD in patients without polyposis syndromes was abdominal pain and vomiting. CONCLUSIONS: Duodenal polyps are most frequently encountered in children with polyposis syndromes, most of whom are asymptomatic. In nonsyndromic patients, the most common histologic subtype is Brunner's gland hyperplastic polyp and presenting symptoms include abdominal pain and vomiting.
Endoscopic findings predict the histologic diagnosis in gastrointestinal graft-versus-host disease.
BACKGROUND AND STUDY AIMS: Graft-versus-host disease (GvHD) of the gastrointestinal tract is a major cause of morbidity and mortality after allogeneic bone marrow transplantation (BMT). Whether endoscopic findings predict the histologic diagnosis of GvHD in the gastrointestinal tract remains controversial. We performed a study to determine the diagnostic accuracy of macroscopic endoscopy findings in the diagnosis of acute and chronic histologically proven gastrointestinal GvHD (GI-GvHD). PATIENTS AND METHODS: Endoscopic images from the intestinal mucosa of post-BMT patients were blindly graded as positive or negative for GI-GvHD and compared with corresponding histological findings, which were used as the gold standard. RESULTS: 44 BMT patients were referred for 96 endoscopic evaluations. Using 162 endoscopy-biopsy pairs, a positive association between endoscopic grading and histologic grading of GI-GvHD (odds ratio [OR] = 11.97, 95% CI 3.86, 37.16) was observed. Endoscopic diagnosis correctly predicted histologic diagnosis in both acute and chronic GI-GvHD (OR = 9.3 vs. 23.1, P = 0.31). CONCLUSIONS: The diagnostic accuracy of endoscopy was high in both acute and chronic histologically proven GI-GvHD. Accurate diagnosis of GI-GvHD might be obtained with mucosal biopsies from either the upper or lower gastrointestinal tract. Endoscopy may play a significant role in establishing early diagnosis and treatment for GI-GvHD in patients following BMT, but histologic evaluation of the gastrointestinal mucosa is needed to confirm the final diagnosis.
Distinctive molecular genetic alterations in sporadic and familial adenomatous polyposis-associated pancreatoblastomas : frequent alterations in the APC/beta-catenin pathway and chromosome 11p.
Pancreatoblastomas are unusual malignant neoplasms of the pediatric pancreas that may also rarely affect adults. The molecular pathogenesis of pancreatoblastomas is unknown. They are clinicopathologically distinct from adult pancreatic ductal adenocarcinomas, but their occasional occurrence in patients with Beckwith-Wiedemann syndrome and the case presented here of a pancreatoblastoma in an adult patient with familial adenomatous polyposis (FAP) suggests that they might bear a genetic similarity to other infantile embryonal tumors such as hepatoblastomas. We analyzed a series of nine pancreatoblastomas for mutations common to other embryonal malignancies including somatic alterations in the adenomatous polyposis coli (APC)/beta-catenin pathway and chromosome 11p, using immunohistochemistry for beta-catenin, 5q and 11p allelic loss assays, and direct DNA sequencing of exon 3 of the beta-catenin gene and the mutation cluster region of the APC gene. In addition, we analyzed the pancreatoblastomas for alterations found in adult-type pancreatic ductal adenocarcinomas including mutations in the K-ras oncogene and the p53 and DPC4 tumor suppressor genes, using direct DNA sequencing of exon 1 of K-ras and immunohistochemistry for p53 and Dpc4. Allelic loss on chromosome 11p was the most common genetic alteration in pancreatoblastomas, present in 86% (six of seven informative cases). Molecular alterations in the APC/beta-catenin pathway were detected in 67% (six of nine), including five neoplasms with activating mutations of the beta-catenin oncogene and the one FAP-associated tumor with biallelic APC inactivation (germline truncating mutation combined with loss of the wild-type allele); seven neoplasms showed abnormal nuclear accumulation of beta-catenin protein. In contrast, loss of Dpc4 protein expression was present in only two cases (one diffuse and one focal), and no alterations in the K-ras gene or p53 expression were detected. Our findings indicate that pancreatoblastomas are genetically distinct from the more common pancreatic ductal adenocarcinomas, but bear a close molecular pathogenesis to hepatoblastomas. In addition, pancreatoblastoma may represent an extracolonic manifestation of FAP.
Sporadic fundic gland polyps: common gastric polyps arising through activating mutations in the beta-catenin gene.
Fundic gland polyps (FGPs) are the most common gastric polyps. FGPs traditionally have been regarded as nondysplastic hamartomatous or hyperplastic lesions, but their pathogenesis remains unclear. We have recently shown that somatic adenomatous polyposis coli (APC) gene alterations are frequently present in FGPs associated with familial adenomatous polyposis (FAP), raising the possibility that mutations of the beta-catenin gene affecting the APC/beta-catenin pathway might be involved in the pathogenesis of sporadic FGPs. We analyzed somatic beta-catenin gene mutations in 57 sporadic FGPs from 40 patients without FAP and in 19 FGPs from 13 FAP patients. Direct DNA sequencing of exon 3 encompassing the glycogen synthase kinase-3beta phosphorylation region for beta-catenin was used with confirmation by HIN:fI restriction endonuclease digestion. The foveolar epithelium and dilated fundic glands of the polyps were separately microdissected and analyzed in 22 of 57 sporadic FGPs. Activating beta-catenin gene mutations were present in 91% (52 of 57) of sporadic FGPs. Both the foveolar epithelium and the dilated fundic gland epithelium comprising the polyps were shown to have the same somatic beta-catenin mutation in 21 of 22 (95%) sporadic FGPs. In contrast, beta-catenin gene mutations were not present in any of the 19 FAP-associated FGPs (P: < 0.000001). The high frequency of beta-catenin mutations in sporadic FGPs indicates that these lesions arise through activating mutations of the beta-catenin gene. Beta-catenin mutations in gastrointestinal tract polyps have previously only been demonstrated in a subset of adenomatous (dysplastic) or neoplastic polyps. Sporadic FGPs are therefore the only lesions of the gastrointestinal tract to demonstrate beta-catenin mutations while lacking dysplastic morphology.
Frequent beta-catenin mutations in juvenile nasopharyngeal angiofibromas.
Juvenile nasopharyngeal angiofibromas (JNAs) are locally aggressive vascular tumors occurring predominantly in adolescent males. The pathogenesis of JNAs is unknown. Recently, JNAs have been reported to occur at increased frequency among patients with familial adenomatous polyposis, suggesting that alterations of the adenomatous polyposis coli (APC)/beta-catenin pathway might also be involved in the pathogenesis of sporadic JNAs. We analyzed somatic beta-catenin and APC gene mutations in 16 sporadic JNAs from nonfamilial adenomatous polyposis patients using immunohistochemistry for beta-catenin, and direct DNA sequencing for exon 3 of the beta-catenin gene and the mutation cluster region of the APC gene. Nuclear accumulation of beta-catenin was diffusely present in the stromal cells but not in the endothelial cells of all 16 JNAs. Activating beta-catenin gene mutations were present in 75% (12 of 16) of JNAs. Six JNA patients also had recurrent tumors after surgery, and in all cases the beta-catenin gene status of the recurrent JNA was identical to the initial tumor. No mutations in the mutation cluster region of the APC gene were detected in the four JNAs without beta-catenin mutations. The high frequency of beta-catenin mutations in sporadic JNAs and the presence of identical beta-catenin gene mutations in recurrent tumors indicates that activating beta-catenin gene mutations are important in the pathogenesis of JNAs. The immunohistochemical localization of beta-catenin only to the nuclei of stromal cells further suggests that the stromal cells, rather than endothelial cells, are the neoplastic cells of JNAs.
Superficial fibromatoses are genetically distinct from deep fibromatoses.
Whereas deep fibromatoses (abdominal, extra-abdominal, mesenteric) display locally aggressive behavior, superficial fibromatoses typically remain small and less likely to recur despite essentially identical morphology. Somatic beta-catenin or APC gene mutations have been reported in < or =74% of sporadic deep fibromatoses and in virtually 100% of Gardner syndrome-associated fibromatoses, whereas genetic events in superficial fibromatoses remain less well characterized. We performed immunohistochemical staining for beta-catenin on 29 superficial fibromatoses (22 palmar, 5 plantar, 1 penile, and 1 infantile digital fibromatosis) and 5 deep fibromatoses. Mutations of beta-catenin and APC genes were analyzed in cases of superficial fibromatoses by direct DNA sequencing of the beta-catenin gene on Exon 3 encompassing the GSK-3 36 phosphorylation region and of the APC gene on the mutation cluster region. Nuclear accumulation of beta-catenin was present in 86% (25/29) of superficial fibromatosis cases ranging from 5 to 100% of nuclei (mean, 13%; median, 10%), though in a minority of nuclei in most examples. Deep fibromatoses had 60 to 100% nuclear staining in all five cases. No somatic mutations of beta-catenin or APC genes were identified in any of the superficial fibromatoses. In contrast to deep fibromatoses, superficial fibromatoses lack beta-catenin and APC gene mutations; the significance of focal nuclear beta-catenin accumulation is unclear. This difference may account inpart for their divergent clinical manifestations despite their morphologic resemblance to deep fibromatoses.
Clinicopathologic and molecular analysis of high-grade dysplasia and early adenocarcinoma in short- versus long-segment Barrett esophagus.
Barrett esophagus, especially dysplastic Barrett mucosa, has been regarded as a preneoplastic lesion for esophageal adenocarcinoma. However, the etiology and pathogenesis of dysplasia and early adenocarcinoma in short- (SSBE) and long- (LSBE) segment Barrett esophagus have not been studied in detail. The aims of this study were to clarify clinicopathologic and genetic differences between high-grade dysplasia (HGD) and early adenocarcinoma in SSBE versus LSBE. We analyzed the clinicopathologic features from 47 patients (19 SSBE [<3 cm] and 28 LSBE [> or =3 cm]) with esophagectomy for HGD/T1 adenocarcinoma. Allelic losses on chromosomes 3p (FIHT), 5q (APC), 9p (p16), and 17p (p53) were compared in 12 HGD and 9 T1 tumors from 19 cases of SSBE and in 23 HGD and 15 T1 tumors from 28 cases of LSBE. Patients with SSBE were more likely to be smokers than were patients with LSBE (94.7% v 57.1%; P =.004). HGD or T1 tumors arising from SSBE were less likely to show adjoining nondysplastic Barrett mucosa than those from LSBE (73.6% v 100%; P =.02). LSBE more frequently showed a circumferential pattern of Barrett mucosa than did SSBE (96.4% v 47.3%; P =.0002). Chromosomal allelic losses on 3p, 5q, 9p, and 17p were detected in 19% (4 of 21), 43% (15 of 35), 40% (14 of 35), and 48% (16 of 33) of HGD, respectively, and 26% (5 of 19), 35% (8 of 23), 35% (8 of 23), and 57% (13 of 23) of T1 tumor, respectively. There were no significant differences in allelic loss of 3p, 5q, 9p, or 17p in HGD or T1 tumors from SSBE versus LSBE. These results suggest that both HGD and early adenocarcinoma in SSBE and LSBE may occur through similar genetic alterations, whereas there are some clinicopathologic differences between SSBE and LSBE. HUM PATHOL
Nasopharyngeal angiofibroma.
Explore the source record for details and available documents.
Hyperplastic polyps of the stomach: associations with histologic patterns of gastritis and gastric atrophy.
Hyperplastic polyps are common gastric lesions characterized by hyperplastic foveolae with variable amounts of inflamed stroma. Their pathogenesis is unknown, but they have been reported to occur in association with various forms of chronic gastritis, particularly autoimmune gastritis and Helicobacter pylori gastritis. Comprehensive histologic evaluation of the background mucosal pathology in patients with hyperplastic polyps has not been previously performed. We studied 160 patients with gastric hyperplastic polyps and characterized endoscopic and histologic features of the polyps (i.e., location, multiplicity, and presence of dysplasia and adenocarcinoma) and the background gastric mucosa (i.e., intestinal metaplasia, dysplasia, carcinoma, and presence and classification of gastritis). Hyperplastic polyps were most common in the antrum (60%) and were multiple in 20% of patients. Focal intestinal metaplasia of the polyp was present in 16% and dysplasia in 4% of patients. Only one patient (0.6%) had adenocarcinoma within the polyp. Evaluation of the surrounding gastric mucosa showed at least focal intestinal metaplasia in 37% of patients, adenoma or low-grade flat epithelial dysplasia in 2%, and synchronous or metachronous adenocarcinoma in 6%. Eighty-five percent of patients had inflammatory mucosal pathology, most commonly active chronic H. pylori gastritis (25%), reactive or chemical gastropathy (21%), and metaplastic atrophic gastritis of the autoimmune (12%) or environmental (8%) type. These results indicate a strong association between various forms of gastritis and the development of hyperplastic polyps and further emphasize the importance of biopsy of the nonpolypoid gastric mucosa during endoscopic examination.
Upper gastrointestinal tract injury in patients receiving kayexalate (sodium polystyrene sulfonate) in sorbitol: clinical, endoscopic, and histopathologic findings.
Kayexalate (sodium polystyrene sulfonate) in sorbitol has been demonstrated to cause colonic necrosis in a subset of uremic patients who are administered the cation exchange resin for treatment of hyperkalemia. Upper gastrointestinal damage associated with Kayexalate in sorbitol is reported far less frequently, and the clinicopathologic spectrum of disease in cases with upper gastrointestinal damage has not been investigated previously. The authors studied the clinical, endoscopic, and histologic features of 11 patients with Kayexalate crystals in biopsies from the esophagus (n = 7), stomach (n = 6), and duodenum (n = 2). The endoscopic appearance was markedly abnormal in all 11 patients. The effects of the medication closely mimicked other endoscopic and radiologic diagnoses in three cases, including esophageal carcinoma, Candidal esophagitis, and gastric bezoar. Histologic and/or endoscopic evidence of mucosal injury in the form of an ulcer or erosion was present in nine patients (82%). In four patients with mucosal injury, no other etiology apart from Kayexalate in sorbitol could be identified. In comparison with a cohort of patients with Kayexalate crystals in lower gastrointestinal specimens identified during the same period (11 patients) the frequency of associated mucosal damage was not significantly different (55%, p = 0.19), but no patient with upper gastrointestinal Kayexalate required surgical resection or died as a result of Kayexalate-induced mucosal injury. The results of this study provide evidence that Kayexalate in sorbitol can induce damage to the upper gastrointestinal tract. Recognition of Kayexalate crystals in histologic sections as a marker for sorbitol-induced mucosal damage may aid in establishing the correct diagnosis for clinically or endoscopically misleading lesions.
Colchicine toxicity: distinct morphologic findings in gastrointestinal biopsies.
Colchicine is an alkaloid with antimitotic ability used to treat a variety of medical conditions. Colchicine toxicity can result in multiorgan failure and death. The histopathologic features of colchicine toxicity in gastrointestinal biopsies have not been reported. Twenty-one gastrointestinal mucosal biopsies obtained from nine patients receiving oral colchicine therapy were studied. Immunohistochemical staining for Ki67 proliferation antigen was performed, and medical records of each patient were reviewed. All patients had a history of gout. Four patients with chronic renal failure also had clinical evidence of colchicine toxicity, and the other five patients did not. Distinct morphologic changes, seen as metaphase mitoses, epithelial pseudostratification, and loss of polarity, were seen in biopsy material from 4 of 4 (100%) patients with clinical colchicine toxicity. Three of these four cases (75%) also contained abundant crypt apoptotic bodies. These morphologic features were best seen in the biopsies from duodenum and gastric antrum, with relative sparing of the gastric body in the upper gastrointestinal tract. Ki67 staining demonstrated an expansion of the proliferating region in three available cases with clinical colchicine toxicity. These distinctive morphologic features were not seen in the five patients without clinical colchicine toxicity. These results indicate that colchicine toxicity can produce diagnostic morphologic features in gastrointestinal mucosal biopsies. Recognition of these features is important because colchicine toxicity can be fatal if undiagnosed clinically.
Hyperplastic polyps of the esophagus and esophagogastric junction: histologic and clinicopathologic findings.
Hyperplastic polyps of the esophagus and esophagogastric junction region (EGJ) are uncommon lesions characterized by hyperplastic epithelium (foveolar-type, squamous, or both) with variable amounts of inflamed stroma. They have been reported almost exclusively in the radiologic and clinical literature as occurring predominantly in association with gastroesophageal reflux disease (GERD). Comprehensive histologic and clinicopathologic evaluation of these polyps, their association with background mucosal pathology, and their association with Barrett's esophagus has not been previously performed. We studied 30 hyperplastic polyps from 27 patients and characterized the histologic, endoscopic, and clinical features of both the polyps and the background esophagus. Hyperplastic polyps were most common in the region of the EGJ (67%), followed by the distal esophagus (30%) and mid-esophagus (3%). Most (80%) were composed of predominantly cardiac-type mucosa, predominantly squamous mucosa (17%), or an admixture (3%). Intestinal metaplasia of the polyp was present in only 7% and low-grade dysplasia in only 3%. In the majority of cases (67%) hyperplastic polyps were associated with concurrent or recent ulcers or erosive esophagitis. In most cases (48%) esophageal injury was associated with GERD, but other potential etiologies included medications, infection, anastomotic or polypectomy sites, vomiting, and photodynamic therapy. Four patients (15%) had Barrett's esophagus, three of whom had or developed dysplastic Barrett's mucosa. These results underscore the pathogenesis of esophageal/EGJ region hyperplastic polyps as a mucosal regenerative response to surrounding mucosal injury. Careful clinical history and biopsy of the nonpolypoid mucosa are essential for determining the clinicopathologic context in which the polyps have developed.
Fundic gland polyps in familial adenomatous polyposis: neoplasms with frequent somatic adenomatous polyposis coli gene alterations.
Fundic gland polyps (FGPs) are the most common gastric polyps in patients with familial adenomatous polyposis (FAP). FGPs have traditionally been regarded as nonneoplastic, possibly hamartomatous lesions, but the pathogenesis of FGPs in both FAP and sporadic patients remains unclear. FGPs in FAP can show foveolar dysplasia, and rarely invasive gastric adenocarcinoma has been reported in patients with FAP and fundic gland polyposis. Using direct gene sequencing and allelic loss assays at 5q, we analyzed somatic adenomatous polyposis coli (APC) gene alterations in 41 FAP-associated FGPs (20 with foveolar dysplasia, six indefinite for dysplasia, and 15 nondysplastic) and 13 sporadic FGPs. The foveolar epithelium and dilated fundic glands of the polyps were separately microdissected and analyzed in 25 of 41 FAP-associated FGPs and 13 of 13 sporadic FGPs. Somatic APC gene alterations were identified frequently (21 of 41 cases, 51%) in FAP-associated FGPs. Both the foveolar epithelium and the dilated fundic gland epithelium comprising the FGPs were shown to carry the same somatic APC gene alteration in 24 (96%) of 25 cases. Furthermore, there was no difference in the frequency of somatic APC gene alterations between FGPs with foveolar dysplasia (10 of 20, 50%), indefinite for dysplasia (four of six, 67%), and nondysplastic (seven of 15, 47%) in FAP patients (P: = 0.697). In contrast, FGPs from non-FAP patients showed infrequent (one of 13, 8%) APC gene alterations (P: = 0.008). These results show that FGPs in FAP patients are pathogenetically distinct from sporadic FGPs. Somatic, second-hit APC gene alterations, which precede morphological dysplasia in many FAP-associated FGPs, indicate that FGPs arising in the setting of FAP are neoplastic lesions.
Esophageal lichen planus: case report and review of the literature.
Involvement of the esophagus by lichen planus is a rarely reported condition. The histologic features of esophageal lichen planus, which may differ from those of cutaneous disease, have only rarely been illustrated. We describe a 58-year-old woman with skin and oral lichen planus who presented with dysphagia and an esophageal stricture that were ultimately diagnosed as esophageal lichen planus. Multiple esophageal biopsies demonstrated a lichenoid, T cell-rich lymphocytic infiltrate, along with degeneration of the basal epithelium and Civatte bodies. Correct diagnosis of esophageal lichen planus is critical because of its prognostic and therapeutic distinction from other more common causes of esophagitis and stricture formation.
Sampling of grossly benign breast reexcisions: a multidisciplinary approach to assessing adequacy.
The widespread use of breast-conserving therapy in the treatment of early-stage breast cancer has resulted in increasing numbers of reexcision specimens requiring histologic assessment for residual disease and margin status. Because many reexcisions are performed for only microscopically positive or close margins, reexcision specimens often appear grossly negative and directed tissue sampling cannot be performed. The issue of adequate sampling in these specimens has not been addressed in the literature. A multidisciplinary approach to identifying the clinically important lesions in breast reexcisions and a cost-effective approach to tissue sampling are needed. We reviewed 97 consecutive cases of grossly negative breast reexcisions in which all tissue had been embedded. Forty-seven specimens contained residual invasive or in situ carcinoma and 50 were histologically negative. Detailed histologic findings were presented to a medical oncologist, a radiation oncologist, and a surgeon, who assessed the clinical impact of each diagnosis. Of the 47 positive specimens, 30 resulted in a major change in patient management (recommendation for additional surgery), 10 resulted in minor changes (alteration in radiation dose or adjuvant chemotherapy regimen), and 7 did not alter management. A total of 1867 blocks were submitted. If one block per centimeter of maximal tissue dimension had been submitted and the remainder of the specimen examined only if initial sections revealed invasive or in situ carcinoma, then 901 blocks would have been processed (52% reduction), but we would have missed an average of 3.7 cases resulting in a major change in therapy, and 3.3 cases resulting in a minor change. In contrast, two blocks per centimeter would have missed an average of less than one case each of diagnoses resulting in major and minor therapy changes (0.9 and 0.8 cases, respectively), and 315 (17%) fewer tissue blocks would have been processed. We recommend submitting two blocks per centimeter in grossly benign reexcisions, and examining the remainder of the tissue only if carcinoma is detected on initial sections.
Erosive injury to the upper gastrointestinal tract in patients receiving iron medication: an underrecognized entity.
Severe gastrointestinal necrosis and strictures after an iron overdose are well described. However, mucosal injury in patients receiving therapeutic iron has received only scant recognition despite its wide use. We studied the clinical and histologic features of 36 upper gastrointestinal tract biopsies from 33 patients (24 gastric, 9 esophageal, 1 gastroesophageal junction, and 2 duodenal) containing characteristic brown crystalline iron material, and evaluated the amount and tissue distribution of the iron. In addition, we investigated the prevalence of iron-associated mucosal injury in upper gastrointestinal endoscopic examinations. The majority of the biopsies (32 of 36, 89%) contained luminal crystalline iron adjacent to the surface epithelium or admixed with luminal fibrinoinflammatory exudate. Thirty biopsies (83%) showed crystalline iron deposition in the lamina propria, either covered by an intact epithelium, subjacent to small superficial erosions, or admixed with granulation tissue. Three biopsies (8%) demonstrated iron-containing thrombi in mucosal blood vessels. Erosive or ulcerative mucosal injury was present in 30 of 36 biopsies (83%). The amount of iron accumulation in cases with mucosal injury was greater than in cases without mucosal injury (mean grades, 2.4+ vs. 1.3+ on a 1+ to 3+ scale; p = 0.002). Iron medication was confirmed in 25 of 33 patients (76%) 22 patients were receiving ferrous sulfate. Approximately half of the patients (17 of 33, 51%) also had underlying infectious, mechanical, toxic, or systemic medical conditions that could have initiated or exacerbated tissue injury. Crystalline iron deposition was found in 0.9% of upper gastrointestinal endoscopic examinations (12 of 1,300), and iron medication-associated erosive mucosal injury was present in 0.7% (9 of 1,300). These results indicate that crystalline iron deposition in the upper gastrointestinal tract is not uncommon. It can induce or exacerbate a distinctive histologic pattern of erosive mucosal injury, especially in patients with associated upper gastrointestinal disorders. Recognition of this pattern by pathologists and its communication to clinicians may aid in optimizing therapy.