The Management of Long-term Parenteral Nutrition
The Management of Long-term Parenteral Nutrition
The amount of functioning small bowel is clearly paramount in determining whether a patient will ultimately require long-term PN support; in general, patients with <75–100 cm of healthy small bowel to an end-enterostomy tend to require long-term parenteral fluid and/or protein-energy support. The presence of a retained colon (jejuno-colonic anastamosis) may allow patients to remain nutritionally autonomous with shorter lengths (sometimes <60 cm) of small intestine. However, the length of residual small bowel only offers a relatively crude prediction of nutritional autonomy and, long-term PN dependency will also, of course, be determined by the presence of intestinal disease. Although it may be difficult to predict exactly whether an individual patient with IF will need long-term PN support, plasma citrulline levels may be a useful biomarker of small bowel mass and function, reflecting absorptive capacity. In an observational study of 57 patients, a plasma citrulline level of <20 μmol/L predicted the development of permanent IF with a sensitivity of 92% and specificity of 90%.
A retrospective European study assessing 124 patients receiving HPN demonstrated that the probability of weaning patients from HPN is <6% if not successfully undertaken in the first 2 years following the last digestive tract modification, ostensibly because the chance of intestinal adaptation thereafter is minimal. Current work evaluating the role of trophic factors aimed at promoting intestinal adaptation that holds some promise; these include both nutrient (e.g. enteral delivery of saturated fatty acids, dietary carbohydrates, glutamine and ornithine) and nonnutrient factors (e.g. growth hormone, epidermal growth factor, insulin-like growth factor, keratinocyte growth factor, leptin and GLP-1 and -2. Growth hormone (Table 1) and GLP-2 (Table 2) have perhaps been studied in the most detail. Growth hormone was first proposed as a therapeutic modality by Byrne in 1995 with an open-label trial. This appeared to demonstrate a benefit to intestinal adaptation in 10 short bowel syndrome patients on long-term PN. The use of growth hormone in intestinal adaptation has been the subject of a Cochrane review, which includes five RCTs containing a total of 79 patients. This meta-analysis suggested a significant increases in weight (mean difference, 1.66; 95% CI, 0.69–2.63), lean body mass (mean difference, 1.93; 95% CI, 0.97–2.9), energy absorption (mean difference, 4.42; 95% CI, 0.26–8.58) and fat absorption (mean difference, 5.02; 95% CI, 0.21–9.82). Adverse events including peripheral oedema (77%), arthralgia (10%) and carpel tunnel syndrome (32%) were reported. Overall, due to the limited numbers of patients assessed in each small RCT, the authors did not feel that there was adequate evidence to support the use of growth hormone for the indication of short bowel syndrome. There is also concern about a potential increased risk of colorectal cancer in patients receiving growth hormone, which may have limited further research.
Perhaps the most promise currently rests with teduglutide, a long-acting GLP-2 analogue has recently received a licence for the treatment of short bowel syndrome from the European medicines agency (Revestive, Nycomed, Zurich, Switzerland) and the Food and Drugs Administration (Gattex, NPS Pharmaceuticals, Bedminster, USA). This has recently been assessed in two multinational double-blind parallel group studies. The first of these phase 3 studies assessed 83 patients on long-term HPN. This demonstrated that 16/35 (46%) patients receiving 0.05 mg/kg/day teduglutide showed a > 20% reduction in parenteral support over 24 weeks compared with 1/16 (6%) patients receiving placebo. Three patients were weaned from parenteral support. Higher doses (0.1 mg/kg/day teduglutide) did not show a significant reduction in parenteral support, although this group did display a trend towards higher baseline parenteral volume, which may have biased the outcome. Teduglutide treatment (0.05 mg/kg/day) had no significant effect on body fat mass, but a modest increase in lean body mass as assessed by DEXA scanning.
A further study then assessed teduglutide at a dose of 0.05 mg/kg/day in 86 patients over 24 weeks with aggressive reductions in parenteral support (10–30%) at two weekly intervals if urine volume increased by more than 10% from baseline. This demonstrated both a statistically significant improvement in the primary end point, a >20% reduction in parenteral support (P = 0.002) as well as an increased plasma citrulline. The mean reduction in parenteral volumes achieved was 4.4 L in teduglutide-treated patients and 2.3 L in placebo-treated patients (P < 0.001).
Patients on long-term PN have been shown to have significantly lower SF36 QoL instrument scores than normal healthy controls. Many patients with IF may never eat or drink again without suffering severe abdominal discomfort and most need to infuse intravenous feed 5–7 nights per week. Thus, while long-term PN may offer many patients a lifeline, not determined dependency can have a detrimental effect on QoL. Enabling home administration of PN therapy and discharge from hospital HPN significantly reduces the cost of care and can allow some patients to return to work. Other factors demonstrating statistically significant effects on QoL include narcotic use, oral fluid volumes, nocturia, the presence of a stoma, age and the number of infusions required per week. Thus, any reduction in the latter that may be afforded by the use of trophic factors will be welcomed.
Retrospective cohorts from large European and North American centres have reported 5-year survival rates between 60% and 78% in unselected patients on PN (Table 3). Survival is principally determined by underlying disease; patients with inflammatory bowel disease for example demonstrate a high 5-year survival of 92%, whereas patients with motility disorders have the poorest 5-year survival at 48%. Multivariate analysis of survival data from single centres has also demonstrated lower survival rates in patients with end-enterostomies or a small bowel length of <50 cm.
The survival of patients receiving PN for advanced malignancy is poor with median time to death of between 5 and 6.5 months. The majority of deaths from HPN (both malignant and nonmalignant) are related to the underlying disease with separate centres reporting only 9% of patients dying of HPN-related complications. Deaths related to the underlying disease tend to occur during the first 2 years of treatment, whereas HPN-related deaths often occur after this.
Outcome of Long-term PN
Weaning From Long-term PN
The amount of functioning small bowel is clearly paramount in determining whether a patient will ultimately require long-term PN support; in general, patients with <75–100 cm of healthy small bowel to an end-enterostomy tend to require long-term parenteral fluid and/or protein-energy support. The presence of a retained colon (jejuno-colonic anastamosis) may allow patients to remain nutritionally autonomous with shorter lengths (sometimes <60 cm) of small intestine. However, the length of residual small bowel only offers a relatively crude prediction of nutritional autonomy and, long-term PN dependency will also, of course, be determined by the presence of intestinal disease. Although it may be difficult to predict exactly whether an individual patient with IF will need long-term PN support, plasma citrulline levels may be a useful biomarker of small bowel mass and function, reflecting absorptive capacity. In an observational study of 57 patients, a plasma citrulline level of <20 μmol/L predicted the development of permanent IF with a sensitivity of 92% and specificity of 90%.
A retrospective European study assessing 124 patients receiving HPN demonstrated that the probability of weaning patients from HPN is <6% if not successfully undertaken in the first 2 years following the last digestive tract modification, ostensibly because the chance of intestinal adaptation thereafter is minimal. Current work evaluating the role of trophic factors aimed at promoting intestinal adaptation that holds some promise; these include both nutrient (e.g. enteral delivery of saturated fatty acids, dietary carbohydrates, glutamine and ornithine) and nonnutrient factors (e.g. growth hormone, epidermal growth factor, insulin-like growth factor, keratinocyte growth factor, leptin and GLP-1 and -2. Growth hormone (Table 1) and GLP-2 (Table 2) have perhaps been studied in the most detail. Growth hormone was first proposed as a therapeutic modality by Byrne in 1995 with an open-label trial. This appeared to demonstrate a benefit to intestinal adaptation in 10 short bowel syndrome patients on long-term PN. The use of growth hormone in intestinal adaptation has been the subject of a Cochrane review, which includes five RCTs containing a total of 79 patients. This meta-analysis suggested a significant increases in weight (mean difference, 1.66; 95% CI, 0.69–2.63), lean body mass (mean difference, 1.93; 95% CI, 0.97–2.9), energy absorption (mean difference, 4.42; 95% CI, 0.26–8.58) and fat absorption (mean difference, 5.02; 95% CI, 0.21–9.82). Adverse events including peripheral oedema (77%), arthralgia (10%) and carpel tunnel syndrome (32%) were reported. Overall, due to the limited numbers of patients assessed in each small RCT, the authors did not feel that there was adequate evidence to support the use of growth hormone for the indication of short bowel syndrome. There is also concern about a potential increased risk of colorectal cancer in patients receiving growth hormone, which may have limited further research.
Perhaps the most promise currently rests with teduglutide, a long-acting GLP-2 analogue has recently received a licence for the treatment of short bowel syndrome from the European medicines agency (Revestive, Nycomed, Zurich, Switzerland) and the Food and Drugs Administration (Gattex, NPS Pharmaceuticals, Bedminster, USA). This has recently been assessed in two multinational double-blind parallel group studies. The first of these phase 3 studies assessed 83 patients on long-term HPN. This demonstrated that 16/35 (46%) patients receiving 0.05 mg/kg/day teduglutide showed a > 20% reduction in parenteral support over 24 weeks compared with 1/16 (6%) patients receiving placebo. Three patients were weaned from parenteral support. Higher doses (0.1 mg/kg/day teduglutide) did not show a significant reduction in parenteral support, although this group did display a trend towards higher baseline parenteral volume, which may have biased the outcome. Teduglutide treatment (0.05 mg/kg/day) had no significant effect on body fat mass, but a modest increase in lean body mass as assessed by DEXA scanning.
A further study then assessed teduglutide at a dose of 0.05 mg/kg/day in 86 patients over 24 weeks with aggressive reductions in parenteral support (10–30%) at two weekly intervals if urine volume increased by more than 10% from baseline. This demonstrated both a statistically significant improvement in the primary end point, a >20% reduction in parenteral support (P = 0.002) as well as an increased plasma citrulline. The mean reduction in parenteral volumes achieved was 4.4 L in teduglutide-treated patients and 2.3 L in placebo-treated patients (P < 0.001).
Quality of Life
Patients on long-term PN have been shown to have significantly lower SF36 QoL instrument scores than normal healthy controls. Many patients with IF may never eat or drink again without suffering severe abdominal discomfort and most need to infuse intravenous feed 5–7 nights per week. Thus, while long-term PN may offer many patients a lifeline, not determined dependency can have a detrimental effect on QoL. Enabling home administration of PN therapy and discharge from hospital HPN significantly reduces the cost of care and can allow some patients to return to work. Other factors demonstrating statistically significant effects on QoL include narcotic use, oral fluid volumes, nocturia, the presence of a stoma, age and the number of infusions required per week. Thus, any reduction in the latter that may be afforded by the use of trophic factors will be welcomed.
Survival
Retrospective cohorts from large European and North American centres have reported 5-year survival rates between 60% and 78% in unselected patients on PN (Table 3). Survival is principally determined by underlying disease; patients with inflammatory bowel disease for example demonstrate a high 5-year survival of 92%, whereas patients with motility disorders have the poorest 5-year survival at 48%. Multivariate analysis of survival data from single centres has also demonstrated lower survival rates in patients with end-enterostomies or a small bowel length of <50 cm.
The survival of patients receiving PN for advanced malignancy is poor with median time to death of between 5 and 6.5 months. The majority of deaths from HPN (both malignant and nonmalignant) are related to the underlying disease with separate centres reporting only 9% of patients dying of HPN-related complications. Deaths related to the underlying disease tend to occur during the first 2 years of treatment, whereas HPN-related deaths often occur after this.
Source...