Total Parenteral Nutrition
Complete dietary replacement with TPN has the longest history of nutrition therapy for IBD, and likely works by eliminating antigenic stimuli from intact food molecules, altering intestinal bacteria, and removing large, nondigestible food components, which generate obstructive symptoms at intestinal strictures. Introduced as primary or adjunctive therapy for severe cases of IBD in the 1960s (
20,
21), early studies were plagued by inconsistent disease activity measurements, clinical end points, and limitations on concomitant steroids (
22,
23). Regardless, TPN appeared to induce CD remission effectively (65% to 100%) but rarely maintained it (0% to 33%) (
24,
25). In addition, long-term TPN can be complicated by line sepsis, access problems, cholestasis, and high cost. TPN results for UC are less impressive, with lower remission rates (27% to 58%) and abysmal maintenance rates (0% to 15%) (
24,
26,
27).
The days heralding TPN as a long-term primary treatment for IBD have passed (
23), because poor efficacy, high cost, and frequent side effects prompt evidencebased guidelines currently to not recommend TPN as a primary treatment for IBD (
28). There may be a role for supplemental parenteral nutrition or TPN as a short-term adjunctive approach to replete micronutrients and provide macronutrients for anabolism when the patient is hospitalized for an acute IBD flare and otherwise consuming minimal or no enteral nutrients orally or cannot tolerate liquid nutrient formulations, but such a practice is not currently evidence based. These circumstances pertain to the management of patients with short gut, prolonged ileus or obstruction, or postoperative fistulas from anastomotic leaks. In these cases, TPN may help to prevent severe malnutrition, particularly in already malnourished patients who may require elective surgery, including further bowel resection resulting from refractory IBD.
Total Enteral Nutrition
In contrast to the foregoing, over time evidence has mounted for the benefits of TEN for the treatment of active IBD. TEN replaces a patient’s caloric and nutrient intake with a processed liquid nutritional supplement administered by mouth or feeding tube. TEN theoretically prevents small bowel mucosal villous atrophy from long-term bowel rest, maintains epithelial integrity, and reduces intestinal immune system activation.
The first randomized, controlled clinical trial demonstrated that TEN could achieve remission as frequently as corticosteroids (
29), but several small, follow-on studies produced conflicting results. A well-executed metaanalysis clarified the situation. For induction of clinical remission, the odds ratio (OR) for steroid therapy versus TEN was 0.35 (95% confidence interval [CI], 0.23 to 0.53), indicating short-term superiority of steroid therapy, but not at 1 year (OR, 0.97; 95% CI, 0.31 to 3.00) (
30). Unfortunately, poor patient compliance with long-term TEN complicates this form of therapy, at least in adult populations.
Compliance rates vary greatly among adult patients with CD from country to country. Although patients in the United States rarely accept long-term TEN, this approach is an important first-line therapy for both inducing and maintaining CD remission in Japan (
31,
32). Similarly, European guidelines promote TEN for adults with active CD complicated by effects of steroids, as supplemental TEN for undernourished children with CD, or as first-line therapy for children with active CD to induce remission (
33).
The percentage of calorie intake from TEN influences success. Consumption of greater than or equal to 900 kcal/ day of TEN resulted in lower hospitalization rates compared with subjects receiving fewer calories (
32). This effect was most noticeable in patients with ileitis. A “half-elemental” diet (half-ED) has also been evaluated for maintaining CD remission (
34). Patients in remission by TEN, prednisolone, induction infliximab, or surgery were randomized to half-ED (900 to 1200 kcal) or free diet (FD) and evaluated for relapse over a 2-year follow-up period. Patients in the half-ED group recurred less frequently (35%) than patients in the FD group (64%; CI = 0.16 to 0.98) (
34).
When examined in aggregate as a primary therapy, TEN induces remission almost as effectively as corticosteroid therapy, without steroid-related side effects. TEN can be administered via tube or mouth, in elemental or semielemental form, and can be continued on a long-term basis to maintain CD remission induced by all standard methods. In pediatrics, EN improves growth velocity hampered by corticosteroid therapy (
35). However, cultural preferences, palatability, feeding tube aversion, and cost hinder widespread use of TEN in the United States.