Crospovidone and Microcrystalline Cellulose: A Novel Description of Pharmaceutical Fillers in the Gastrointestinal Tract.

Crospovidone and microcrystalline cellulose (MCC) are pharmaceutical fillers well known in the pulmonary pathology literature. Fillers are inactive substances incorporated into medications to facilitate drug delivery. By examining 545 consecutive gastrointestinal surgical specimens from 302 patients between September 11, 2015 and October 23, 2015, we identified the fillers in 29 specimens from 26 patients. The control group consisted of an equal number of consecutive site-matched specimens collected during this same time. Pertinent clinicopathologic data were analyzed, and 1 case was subject to special stains. To confirm the histologic diagnosis, a variety of fillers and medications common to the patients were processed. The fillers were found in 9% of all patients, and there were no specific clinicopathologic associations. In the gastrointestinal tract, crospovidone is nonbirefringent and has a coral shape with each segment composed of a pink core and purple coat; MCC is brightly birefringent with matchstick shape and clear color. Identical material was seen in the processed crospovidone and MCC powders, as well as oxycodone-acetaminophen and omeprazole tablets. In summary, crospovidone and MCC are common, biologically inert, and they are most often seen in the small bowel. Their presence outside of the luminal bowel may serve as a surrogate marker for perforation. Awareness of their morphology is important to distinguish fillers from parasites, calcifications, and other medications, particularly those linked to mucosal injury. We report the unique histomorphologic profile of these fillers as a helpful diagnostic aide, and caution that the fillers have slightly divergent features when compared with those described in the lung.

Neutropenic Enterocolitis: New Insights Into a Deadly Entity.

Neutropenic enterocolitis (NE) is a deadly ileocecal-based disease seen in patients with a recent history of chemotherapy. As histology is not included in the current diagnostic criteria, the pathologic features of NE are poorly understood. We undertook a multi-institutional study of NE, and report helpful clinical clues, such as immunosuppression (n=20/20), recent chemotherapy (n=17/18), neutropenia (n=16/18) gastrointestinal symptoms (n=19/19), abnormal imaging studies of the cecum/right colon (n=11/14), and positive microbiological studies (n=13/15). Fever (n=9/15) and sepsis (n=8/16) were also common. Pathologically, the cecum/right colon was always involved (n=17/17), but findings were identified in other bowel segments as well. NE lesions consisted of patchy necrosis (n=18/20), infiltrating organisms (n=17/20), hemorrhage (n=15/20), ulcer (n=15/19), edema (n=15/20), and depletion of inflammatory cells (n=15/20). Seventy-nine percent (n=15/19) of patients with histologically confirmed NE died: 47% (n=7/15) of these deaths were attributed to NE and the remainder to the patients' underlying conditions. Importantly, we observed a clinical diagnostic discordancy rate of 35% (n=9/26): 15% (n=3/20) of histologically confirmed NE were clinically unsuspected, and 26% (n=6/23) of clinically suspected NE represented a different disease process. Alternative diagnoses included unspecified colitis, infection, graft-versus-host disease, relapsed malignancy, mycophenolate injury, appendicitis, and ischemia. The causes of death in patients with NE mimics included unrecognized appendicitis and unrecognized graft-versus-host disease. To improve diagnostic accuracy, we propose that histology be required for a diagnosis of "definitive NE," with other clinically suspicious cases reported as "suspicious for NE" until all other possible diagnoses have been reasonably excluded.

Colesevelam and colestipol: novel medication resins in the gastrointestinal tract.

We report the morphologic description of the bile acid sequestrants (BAS) colesevelam and colestipol, as well as the largest series of cholestyramine. Histologically similar medication resins from 4 institutions were prospectively collected over 1 year (26 specimens, 15 patients). Comorbidities included hyperlipidemia (4/15), hypertension (4/15), inflammatory bowel disease (4/15), coronary artery disease (3/15), diarrhea (7/15), hypothyroidism (2/15), and ischemic bowel (1/15). Sites of involvement included the esophagus (1/26), stomach (1/26), small intestine (1/26), ileocecal valve (1/26), and colorectum (22/26). Associated histologic diagnoses included normal (8/26), chronic mucosal injury (11/26), acute inflammation (9/26), erosion/ulceration (6/26), and cytomegalovirus (2/26). The BAS resins were histologically indistinguishable from each other; they were all eosinophilic on hematoxylin and eosin (H&E) and lacked internal "fish-scales." To validate these observations, respective medications were submitted for histologic processing; the processed medications were identical to those in the patient specimens. Rare, irregular "fracture" lines presented diagnostic pitfalls by mimicking the true "fish-scales" of Kayexalate and sevelamer. Clues to the correct identification of BAS include recognition that the "fracture" lines were subtle, irregular, and restricted to large fragments or thick sections, likely representing a processing artifact. Moreover, Kayexalate is violet on H&E and black on acid fast bacillus, and sevelamer characteristically displays a 2-tone color on H&E and is magenta on acid fast bacillus. An association with inflammatory injury was seen (15/26). We believe that the BAS are innocent bystanders in complicated patients, although we cannot exclude their ability to cause mucosal injury in specific settings.