Neutropenic Enterocolitis: Clinical Features and Outcomes

Background. Little attention has been given to patients with neutropenic enterocolitis (NEC or typhlitis). However, as significant advances have been made in both hematology, oncology and critical care management, updating outcome data on these patients is timely.

Patients and methods. We conducted here a retrospective hospital-based cohort from patients admitted to the intensive care unit (ICU) from a cancer center that includes 10 wards of oncology and hematology (a 650-bed public hospital with 330 beds for treatment of malignancies). Patients with NEC were compared to patients with other digestive complications. We used the NEC definition suggested by Gorschlüter based on presence of neutropenia and at least one major criteria (bowel thickening and fever) with one or several minor criteria (abdominal aches, diarrhea, abdominal distension, abdominal cramps and low digestive bleedings), after exclusion of other intestinal affections listed above.

Results. Among the 171 patients admitted for acute digestive complications over a 15-y study period (1997-2011), 54 (33%) presented with NEC. Median age was 48 years (IQR: 38-60y) and there were 34 men and 20 women. Hematological malignancies were more common (87% versus 13% solid tumors), mainly non hodgkin lymphoma (41%) and among them 5 undergoing autologous SCT. Abdominal pains were present in 74% of the patients. Vomiting, diarrhea and mucositis were present in 91%, 76% and 39%, respectively. All patients were treated with fasting, naso-gastric aspiration, hydroelectrolytic equilibration, large spectrum antibiotics and parenteral nutrition. Eight (15%) patients were operated. Vasopressive drugs were used for 31 patients (57%), mechanical ventilation for 22 (41%) and renal replacement for 14 (26%). Microbiological infection was documented for 32 patients: 1 fungal infection (Candida), 16 Escherichia coli, 7 Clostridium difficile (2 toxinogen), 6 Pseudomonas aeruginosae, 2 other enterobacteria. Twelve (22%) patients died.

Compared to non-NEC patients, patients with NEC were sickest (median Simplified Acute Physiology II score was 54.5 (IQR 42.5-62.5) vs. 48 (IQR 35.5-61.5). Also, NEC was associated with more exposure to cytosine arabinoside (46% vs 20%, p=0.0008), anthracyclines (37% vs 13%, p=0.0008), vinca-alkaloids (28% vs 8%, p=0.002) and VP16 (24% vs 12%, p=0.0067). Steroids were not associated with incidence of NEC (19% vs 21%, p=0.84). Clinical symptoms that were significantly different in patients with NEC were vomiting (91% vs 70%, p=0.003), diarrhea (76% vs 53%, p=0.006) and mucositis (39% vs 17%, p=0.004). However, digestive bleedings were less common (11% vs 30%, p=0.01). CT findings helped to differentiate NEC and non-NEC patients with, colic (57% vs 33%, p=0.033) and small intestine (35% vs 14%, p=0.023) thickening, occlusive findings (51% vs 19%, p=0.002). Pneumatosis and pneumoperitoneum were not significantly different in the two groups (8% vs 4%, and 8% vs 12%, respectively). Need for surgery was non significantly higher in non-NEC patients (25% vs. 15%, p=0.22). However, mortality was significantly lower in NEC patients (22% vs 41%, p=0.023). Median time hospitalization in ICU was 7 days (IQR: 6.5-23.5d) vs 4d (IQR: 3-10d) for non-NEC patients.

Conclusion: Neutropenic enterocolitis is a clinical entity that has its own clinical, radiological and management peculiarities that translate into different outcomes. Early recognition of the syndrome may help in setting prompt appropriate treatment and monitoring.