Primary CMV Colitis in an Immunocompetent Infant, Successfully Treated by Gancyclovir

Le résumé fourni par la source

Cytomegalovirus (CMV) is extremely rare as a cause of gastrointestinal infection in the immunocompetent child, despite the high seroprevalence rate of CMV in the general population (1,2). Where isolated cases have been reported, symptoms have resolved spontaneously with time (3). We report the case of a 12-week-old immunocompetent infant who was admitted acutely with a 10-day history of diarrhea, abdominal pain, and high fever; investigations suggested inflammatory bowel disease, but histological examination of colonic biopsies showed typical features of CMV infection. When severe symptoms persisted gancyclovir treatment was followed by a full recovery, sustained during a 5-year follow-up. CASE REPORT A 12-week-old fully immunized breast-fed infant girl was admitted with a 10-day history of passing liquid diarrhea with mucus every 2 to 3 hours, fever, abdominal discomfort, and poor feeding. Her temperature on admission was 38.5°C, and she looked strikingly pale. Examination revealed a well-nourished child with a slightly distended abdomen, but no tenderness on palpation. Her weight was 6.11 kg, which was consistent with her birth weight of 3.7 kg. Initial investigations showed her hemoglobin to be 65 g/L, platelet count 707 × 109/L, reticulocyte count 3%, C-reactive protein 35 g/L, albumin 31 g/L, and bilirubin 4 μmol/L. Ultrasonography of the abdomen showed that the descending colon had a thickened appearance. After receiving a blood transfusion, she became less miserable and her feeding improved. Over the next few days, there was a reduction in abdominal discomfort, and the diarrhea and fever subsided. Cultures of the stool, urine, and blood were negative. She was discharged from the hospital with the diagnosis of probable infective gastroenteritis, with the plan that further investigations including colonoscopy would be undertaken if her symptoms recurred. At an outpatient review 1 week later, she remained well and free of fever, but after a further week the diarrhea returned, with mucus and flecks of blood. She cried with pain with the passage of each stool, and her weight had dropped slightly from 6.33 to 6.12 kg in 1 week. Investigations showed her hemoglobin to be 108 g/L, C-reactive protein 24 g/L, platelets 538 × 109/L, and serum sodium 129 mmol/L. She was prescribed sodium supplements (2 mmol · kg−1 · day−1), and cow's milk protein was excluded from her mother's diet in case the diarrhea was related to cow's milk protein intolerance. Colonoscopy confirmed a marked distal colitis extending from the rectum to the splenic flexure. Histologically, this seemed typical of ulcerative colitis, but “owl's eye” inclusions were also seen, indicating the presence of CMV in the mucosa (Fig. 1); CMV was also seen in a urine specimen. At this stage we were concerned about the possibilities of underlying inflammatory bowel disease, complicated by CMV infection, and underlying immunodeficiency.FIG. 1: Colonic mucosal biopsy specimen showing inflammatory infiltrate and inclusion body typical of CMV infection (arrow). (Hematoxylin-eosin; original magnification × 400.)A full immunological assessment was performed, including the following tests: T cells (CD3), B cells (CD19), T helper cells (CD4), T suppressor cells (CD8), CD4/CD8 ratio, and natural killer cells (CD16+56), all of which were normal. She continued to be unwell, and 9 weeks after initial presentation she was still experiencing bloody diarrhea, abdominal pain, irritability, poor feeding, weight loss, and spiking fever. Her inflammatory markers remained elevated, the albumin was low (23 g/L) and her hemoglobin had dropped to 66 g/L. A central venous catheter was inserted, and treatment was given initially with 2 infusions of immunoglobulin 1 week apart. When this had little effect, treatment with intravenous gancyclovir was commenced at a dosage of 5 mg/kg twice daily; this dosage was halved after 48 hours when she experienced a neutropenia and continued for a further 12 days. Within 3 days after the beginning of antiviral treatment, she showed a clinical response, with settling of fever, reduced abdominal pain, and diminution in diarrhea. This progress continued for the next 2 weeks until the fever, pain on defecation, and diarrhea completely resolved; her feeding improved; and catch-up weight gain was observed. Four weeks after discharge, her weight had increased to 7.34 kg and her length to 68.3 cm. Her bowels were opening 4 or 5 times each day; the stools were soft and without blood or mucus. She remained breast-fed and consumed a milk-free diet until 1 year of age, and was readmitted to hospital with pneumococcal pneumonia at age 18 months. She made a rapid recovery with intravenous penicillin; tests of immune function were repeated, and once again the results were normal. At 5 years of age she remains well with normal growth and bowel habit, and has recently had an uncomplicated episode of chickenpox. DISCUSSION As many as 50% of individuals have had a CMV (a herpes group DNA virus) infection by adult life; the infection is usually asymptomatic. Transmission may be from urine, saliva, blood, or breast milk, and the incubation period is from 3 weeks to 3 months. CMV is an important cause of morbidity and mortality in immunosuppressed patients such as those with AIDS, recipients of organ transplants, burn patients, and those receiving immunosuppressive therapies (4–6). CMV infection may also cause an exacerbation of preexisting inflammatory bowel disease (4). Congenital CMV infection usually follows asymptomatic maternal infection, and only 5% of infected infants have an acute infection at birth (4,7). The diagnosis of congenital CMV infection is based on cultures from urine or the throat within the first 2 weeks of life. After this time, if polymerase chain reaction results are positive and the mother's CMV status is negative, it is more likely to be an acquired infection. Classic CMV inclusions may be difficult to see on histological examination of colonic mucosa, and CMV culture may be negative even when inclusions are seen (positive in 50% to 80%) (4,6,8). The mucosa may have a normal appearance at endoscopy despite CMV infection on histology. CMV DNA polymerase chain reaction on blood may be negative because viremia is intermittent (4,9). CMV IgM is usually positive by 3 to 4 weeks of life in congenital infection, and it becomes positive at 8 to 12 weeks of age in perinatal or postnatal infection (4,10,11). We initially considered the possibility of an allergic colitis in our patient; however a response to milk protein exclusion was absent, and the typical histological appearances of eosinophils in the mucosa were absent. CMV infection in association with allergic colitis has been described in an infant, but symptoms resolved with dietary exclusion (12). Macroscopic and histological findings from the colon were suggestive of ulcerative colitis in our patient, but her rapid response to gancyclovir and her subsequent clinical course make this diagnosis very unlikely. Oral treatment may have been an option, but there was little published information about the effectiveness of this treatment for active CMV infection, and we felt justified in using intravenous gancyclovir, given her continuing ill health and severe symptoms, including diarrhea. Oral treatment provides much lower tissue levels than intravenous administration and is mainly restricted to prophylaxis against CMV disease in those at risk (13). This was done after careful discussion with the family, even though this agent (an analogue of guanine, which inhibits replication of herpesviruses) has been associated with carcinogenicity and reproductive toxicity when used over the long term. About two thirds of patients will have negative antigenemia at the end of therapy, CMV antigenemia will remain positive in 20%, and about one third will need a second course of treatment (14). Neutropenia and fever are common side effects; others incl

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.