Mj. Page et al., CYTOMEGALOVIRUS ENTERITIS - A HIGHLY LETHAL CONDITION REQUIRING EARLYDETECTION AND INTERVENTION, Diseases of the colon & rectum, 41(5), 1998, pp. 619-623
Cytomegalovirus infection causing symptomatic enteritis is most usuall
y associated with immunosuppressed transplant patients or patients pos
itive for human immunodeficiency virus. Most reports studying this ill
ness are small and do not clearly define the risk factors or mortality
rates. METHODS: The present study retrospectively reviewed the charts
of 67 patients with biopsy-proven cytomegalovirus enteritis (esophage
al, gastric, small bowel, and colonic) to define and to investigate fa
ctors that influence survival. Patients were classified into four grou
ps based on underlying medical condition: 1) patients positive for hum
an immunodeficiency virus; 2) transplant patients receiving immunosupp
ressive medications; 3) immunosuppressed nontransplant patients; and 4
) otherwise healthy individuals. Mortality rates based on underlying m
edical condition, location of intestinal cytomegalovirus infection, cy
tomegalovirus therapy, age, and average days to institution of treatme
nt were defined and statistically assessed. RESULTS: Mortality was sig
nificantly greater in the normal patient group (80 percent) than in th
e transplant (21 percent), other immunosuppressed (44 percent), or hum
an immunodeficiency virus-positive (75 percent) groups (P = 0.0006, Co
chran-Mantel-Haenszel statistics). There was no difference in mortalit
y based on intestinal location of disease or treatment modality (surge
ry, medical therapy, or both). Cohorts of patients order than 65 years
had a statistically higher mortality rate vs. those younger than 65 y
ears old (68 vs. 38 percent; P = 0.05, Cochran-Mantel-Haenszel statist
ics). Statistically increased mortality was also associated with incre
ased time from hospital admission to institution of cytomegalovirus tr
eatment, whether therapy was medication alone or medication and surger
y (P < 0.05, exact Wilcoxon's test). CONCLUSIONS: 1) Lethal cytomegalo
virus enteritis can arise in patient populations not typically identif
ied as being at risk for this disorder, including normal individuals.
2) Mortality in cytomegalovirus enteritis is adversely associated with
age older than 65 years and increased time to institution of therapy
but is not affected by anatomic site of infection or particular form o
f treatment. Paradoxically, in this study, normal patients had the hig
hest mortality, which we attribute to a low index of suspicion acid re
latively late institution of therapy.