Ulcerative proctitis is defined as a chronic inflammatory process limited to the rectum. It is generally considered to represent one clinical variant of ulcerative colitis, and in studies of adults constitutes ≈20-35% of newly diagnosed cases of ulcerative colitis (1-4). In contrast to more extensive ulcerative colitis, ulcerative proctitis is thought to follow a more benign course with less severe symptoms and a decreased propensity to develop cancer (5). Studies of adults have suggested that proximal extension with more significant colonic involvement may occur in 10-30% of subjects, and may occur either early or late after diagnosis (5,6).
Data on ulcerative proctitis in children are limited. In one study of ulcerative colitis in children with onset of symptoms at 10 years of age or younger, disease was limited to the rectum in 6% of patients (7). There is one previous report of ulcerative proctosigmoiditis in children that included 85 patients whose disease onset was before 21 years of age (8). Proximal extension of disease and more severe symptoms occurred in ≈40% of subjects, and colectomy was eventually required in ≈15%. To better characterize the frequency, clinical presentation, and outcome of ulcerative proctitis in children, we have conducted a review of the clinical experience with this disorder at several medical centers.
A review was conducted of the case records of all patients with a diagnosis of inflammatory bowel disease (1975-1994) from the Divisions of Pediatric Gastroenterology and Nutrition at the following institutions: Hartford Hospital, Hartford, CT; University of Vermont College of Medicine, Burlington, VT; Children's Hospital, Boston, MA; Mt. Sinai Hospital, New York, NY; and North Shore University Hospital, Manhasset, NY. For inclusion in the study, a patient needed to meet standardized criteria for ulcerative proctitis and have a minimum duration of follow-up of 1 year.
Ulcerative proctitis was defined endoscopically as rectal mucosal inflammation extending up to but not beyond 15 cm proximal to the dentate line, normal-appearing mucosa proximal to the “edge” of inflammation, and histology compatible with idiopathic proctitis including distored crypt architecture or other features of chronicity, lamina propria inflammation, crypt abscess, and neutrophils in the surface epithelium. The presence of a granuloma on rectal biopsy or perianal disease such as fissure or tags was considered suggestive of Crohn's disease and excluded a diagnosis of ulcerative proctitis. Appropriate stool cultures excluded identifiable enteric pathogens as a cause of symptoms. Grossly and histologically normal mucosa proximal to 15 cm from the dentate line was confirmed using flexible sigmoidoscopy or colonoscopy.
Disease Severity at Presentation
Mild disease was defined as the intermittent presence of blood generally accompanied by four or fewer stools per day. Moderate or severe disease was defined as the daily presence of bleeding with the frequent passage of five or more stools per day. An attempt was made to assign each patient to one of these groups based on their reported symptoms during the several weeks before diagnosis.
Disease Activity During Course
Asymptomatic or inactive disease was defined by the absence of bleeding or diarrhea and no systemic symptoms. A chronic intermittent course was defined by the occurrence of bleeding-free intervals of at least 1-month duration, excluding periods when the patient was taking corticosteroids orally. A continuous course meant there were no bleeding-free intervals of at least 1-month duration, or that the patient required daily corticosteroid therapy (at least 50% of the time) to suppress symptoms and that symptoms recurred when the corticosteroids were discontinued.
Given the multicenter nature of this retrospective review, there was no defined treatment protocol applied to all patients. However, local therapy was generally used first and included mesalamine suppositories or enemas and corticosteroid suppositories, enemas, or foam. Some patients were also treated orally with sulfasalazine, mesalamine, and/or systemic corticosteroids.
Differences between groups were evaluated using contingency table analysis (χ2 statistics and Fisher's Exact Test, when indicated) for categorical variables and t tests for continuously distributed variables. Kaplan-Meier product-limit estimates were used to describe the probability of disease extension over time.
A total of 38 subjects (22 female) were identified, representing ≈6% of all newly diagnosed cases of ulcerative colitis seen at the participating centers during the study period. The characteristics of these subjects are shown in Table 1. They ranged in age from 4.2 to 17.7 years (mean ± SD age 11.6 ± 3.9 years). Forty percent were 10 years of age or younger at diagnosis. Sixteen percent had a family history of inflammatory bowel disease.
Sixty-three percent had symptoms for <3 months before diagnosis. Mild symptoms were present at diagnosis in 74%, with the remainder being more severe. Twelve patients (32%) had a complaint of constipation at the time of diagnosis. Duration of follow-up for the study patients ranged from 1-7 years (mean 4.3 ± 3.4 years).
Arthritis was noted at the time of diagnosis in three patients and developed during the subsequent course in two others. One subject developed erythema nodosum subsequent to diagnosis of proctitis. No other extraintestinal manifestations of inflammatory bowel disease were noted.
Laboratory studies at diagnosis included hematocrit ≥35% in 31 of 36 patients, with the lowest presenting hematocrit being 31%; erythrocyte sedimentation rate <20 mm/h in 26 of 32 patients, with the elevated rates ranging from 20 to 27 mm/h, and albumin ≥3.7 g/dl in 26 of 26 patients.
Initial Response to Therapy
Specific treatment modalities used in the 1st month of therapy included rectal 5-aminosalicylates (12 of 38 patients), rectal corticosteroids (13 of 38 patients), oral 5-aminosalicylates (22 of 38 patients), and oral corticosteroids (1 of 38 patients). One additional subject was started on oral corticosteroids between 3 and 6 months after diagnosis. Cessation of symptoms was observed within 3 months in 68% of patients and in 3-6 months in an additional 24%. Three patients (8%) still had symptoms 6 months after diagnosis. All three of these persistently symptomatic children had been classified as having mild disease at presentation.
For yearly follow-up intervals starting at 6 months postdiagnosis, on average ≈55% of patients available for follow-up were asymptomatic, 40% had a chronic-intermittent course, and <5% were continuously symptomatic despite therapy. Oral corticosteroids were eventually used in an additional six patients between 6 months and 5 years after diagnosis. Two of these patients were started on corticosteroids for continuous symptoms and four for chronic-intermittent disease. Two subjects received immunomodulator therapy when they developed proximal extension of disease, which was unresponsive to corticosteroid therapy.
We examined the data to see whether we could define clinical or laboratory factors that might be predictive of relapse. Of the 35 subjects entering remission within the first 6 months of therapy, 19 sustained remission in the second 6 months and 16 relapsed. There was no significant difference in the two groups with respect to sex or age at diagnosis, initial laboratory parameters, whether rectal medications were used, severity of initial disease, or duration between institution of initial therapy and remission after diagnosis.
Extension of inflammation proximal to the rectosigmoid occurred in 11 of the 38 study subjects (29%) 0.5-11.3 years after diagnosis; 7 demonstrated extension within 3 years of diagnosis. The mean ± SD and median durations from diagnosis to extension in these 11 study subjects were 4.1 ± 3.6 years and 2.6 years, respectively. Inflammation involved the descending colon in eight and the entire colon in three. Ten of these 11 patients had been classified as having mild disease at presentation.
The Kaplan-Meier product-limit estimate of the likelihood of disease extension is ≈25% by 3 years (22 patients were observed for a minimum of 3 years). Because of limited longer follow-up, probabilities of extension beyond 5 years are unstable. In the 13 patients for whom a minimum follow-up of 5 years was available, 3 had extension in the first 5 years of their course, and 4 had extension at 6, 7.5, 8, and 11 years after diagnosis, respectively. There was no statistically significant relationship between initial severity of symptoms, laboratory studies at presentation, age at diagnosis, or use of rectal therapy and the likelihood of developing proximal extension of disease.
Two subjects (5% of all study patients) have undergone colectomy. One patient was 8.8 years of age at diagnosis of mild proctitis, had proximal extension of disease at age 14 years, and had colectomy 6 months later. The other patient was 6.8 years of age at diagnosis, had extension to pancolitis 6 months later, and required colectomy despite therapy with cyclosporine.
In the current study, we have characterized the clinical course of 38 pediatric subjects with ulcerative proctitis. All had a minimum follow-up of at least 1 year, with a mean duration of follow-up of >4 years. Although many study subjects followed a benign course, we also found a high likelihood of proximal extension of disease. This finding is similar to that found by Mir-Madjlessi et al. (8) in their report of an older pediatric population (85 subjects, mean age 16 years) with proctosigmoiditis diagnosed between 1958 and 1983, a period in which the newer 5-aminosalicylic agents were not yet available. In the 66 subjects in their series who had evaluation that would allow determination of proximal extension of disease, inflammation was eventually noted to involve the descending colon in 13 and the more proximal colon in 25 (58% of the subjects had extension beyond the rectosigmoid). Although the timing of disease extension was variable, it generally occurred within 5 years of onset of symptoms. In the present study, proximal extension was noted by 3 years in 25% of subjects, but extension developed as long as 11 years after diagnosis. In studies of adults, the rate of proximal extension has ranged from 10 to 30% (5,6).
We were not able to identify any specific laboratory or clinical features that identified those subjects at greatest risk of proximal extension of disease. Hematocrit and erythrocyte sedimentation rate were within normal limits in >80% of subjects, and serum albumin levels were normal in all of those who had this determination done. There was a suggestion that lack of early response to initial therapy or early relapse after initial remission might identify subgroups of patients at greater risk. However, the relatively small number of patients studied suggests that the data in this regard need to be interpreted with caution.
Extraintestinal manifestations were recorded in six patients during their course, including three in whom arthritis was present at diagnosis. These data are similar to those reported in adolescents with proctosigmoiditis (8).
Unfortunately, the retrospective nature of our study does not allow evaluation of the efficacy of specific therapies for proctitis. Many subjects were treated with more than one modality including both oral as well as rectal medications. Sixty-eight percent of subjects had no symptoms by 3 months and >90% by 6 months. A total of eight patients were treated with oral corticosteroids during their course.
Characterization of the clinical pattern beyond 6 months from diagnosis as asymptomatic, chronic intermittent symptoms, or continuous symptoms was similar for our group of subjects to that recently noted in 171 children with ulcerative colitis (9). We found that, on average, ≈55% of our subjects were asymptomatic during any yearly follow-up interval, ≈40% had chronic intermittent symptoms, and <5% were continuously symptomatic. One of the patients in the present study required colectomy within the first 5 years postdiagnosis, and one additional subject required colectomy at 6 years. In contrast, in the group of children with ulcerative colitis, 19% required colectomy within 5 years (9). Given the size of our patient population and the nonstandardization of treatment protocols, we were not able to examine the possibility that either continuous administration of oral 5-aminosalicylate agents or rectal preparations may have affected a patient's course.
In conclusion, our study suggests that counseling of children and their families regarding the prognosis of ulcerative proctitis should include information on the likelihood of proximal extension of disease, and the occasional need for colectomy. The identification of children with proximal extension of disease will identify a group of subjects who will require endoscopic surveillance for dysplasia when their disease course lasts more than 8-10 years. Prospective multicenter studies will be required to evaluate the efficacy of treatment protocols to treat ulcerative proctitis, and possibly decrease the likelihood of proximal extension.