Umbrella review of the efficacy of perioperative immunonutrition in visceral surgery.

INTRODUCTION: Immunonutrition (IN) is used in major visceral surgery to reduce postoperative complications. This umbrella review (review of reviews) collects and analyses data on the efficacy of perioperative IN. METHODS: The review was conducted in accordance with PRISMA 2020 guidelines. Inclusion criteria were meta-analyses comparing IN with normal diet or isocaloric isonitrogenous feeding. The primary outcome was infectious complications. Secondary outcomes were overall morbidity, hospital length of stay and mortality. Methodological quality was evaluated using AMSTAR-2. Overlap and certainty of evidence (GRADE) were assessed. RESULTS: Twenty meta-analyses (MAs) were included in the umbrella review: eleven on various abdominal surgeries (one MA was considered twice) and eight on pancreatic, oesophageal, hepatic, or colorectal surgeries. Overall, IN was associated with significantly fewer postoperative infectious complications (OR 0.60 [0.54-0.65], random effect model) but with substantial heterogeneity (I2 = 64%), and less postoperative morbidity (OR 0.78 [0.74-0.81], I2 = 30.3%). Excluding three MAs with heterogeneity did not alter the results. The overlap between the MAs was slight, with a corrected covered area of 0.13. There was no significant difference in the timing of IN (preoperative, postoperative or perioperative). CONCLUSION: This umbrella review confirms the beneficial effect of IN in visceral surgery. Some practical questions remain unanswered: optimal timing of IN, in which surgical speciality it is best used, and its utility in enhanced recovery programmes. REGISTRATION IN PROSPERO: CRD42021255177.

[Refeeding syndrome]. / Syndrome de renutrition inappropriée.

Refeeding syndrome (RS) is a rare but severe condition that is poorly understood, often under-diagnosed and can lead to death. It occurs within 5 days after refeeding in patients after prolonged fasting or in a context of undernutrition. As a consequence of the abrupt transition from catabolism to anabolism, RS is defined as a decrease in plasma levels of phosphorus, potassium and/or magnesium, whether or not associated with organ dysfunction resulting from a decrease in one of the electrolytes or a thiamine deficiency, after refeeding. The clinical symptoms are varied and non-specific and are related to hydro electrolyte disorders, sodium-hydroxide retention or failure of one or more organs. Patient management should be appropriate with regular clinical examination and careful biological monitoring, including hydro electrolyte monitoring. The correction of hydroelectrolytic disorders and systematic thiamine supplementation are essential during refeeding, that must be done carefully and very progressively, whatever its form (oral, enteral or parenteral). The severity of the refeeding syndrome indicates that its prevention and screening are the corners of its management in at-risk patients.

Six-month outcomes of teduglutide treatment in adult patients with short bowel syndrome with chronic intestinal failure: A real-world French observational cohort study.

BACKGROUND & AIMS: Teduglutide, a GLP-2-analog, has proven effective in two placebo-controlled studies in reducing parenteral support (PS) in patients with short bowel syndrome-associated intestinal failure (SBS-IF) after 24 weeks. The aim of this study was to describe in a real-life situation the effects of teduglutide treatment and their predictive factors. METHODS: We included 54 consecutive SBS-IF patients treated with teduglutide in France for at least 6 months from 10 expert centers. Small bowel length was 62 ± 6 cm and 65% had colon in continuity. PS was 4.4 ±0 .2 infusions per week, started 9.8 ± 1.2 years before. Response (PS reduction ≥ 20%) and PS discontinuation rates were assessed at week 24. Adjusted p values of factors associated with response and weaning were calculated using a multivariate logistic regression model. RESULTS: At week 24, 85% of patients were responders and 24% had been weaned off PS, with a 51% reduction of PS needs and 1.5 ± 0.2 days off PS per week. Response to teduglutide was influenced by a higher baseline oral intake (p = 0.02). Weaning off PS was influenced by the presence of colon (p = 0.04), a lower PS volume (p = 0.03) and a higher oral intake (p = 0.01). There were no differences based on age, bowel length or SBS-IF causes. CONCLUSIONS: Our study confirms the effectiveness of teduglutide in reducing PS needs in SBS-IF patients. We associated reduced parenteral support volume with baseline parenteral volume support, bowel anatomy, and oral intake. These findings underline the role of nutritional optimization when starting the treatment.

Home parenteral nutrition (HPN) in patients with post-bariatric surgery complications.

BACKGROUND & AIMS: Obesity is a worldwide health problem. Bariatric surgery (BS) is becoming one of the most commonly used methods for fighting obesity and its associated comorbidities. However, current BS techniques can be associated with early or late complications that may require nutritional support. The aim of this retrospective observational study was to determine the indications and outcomes for patients on Home parenteral nutrition (HPN) due to post-bariatric surgery complications. METHODS: A specific questionnaire was designed by the ESPEN HAN/CIF working group and submitted to HPN centers. This questionnaire included: patient demographics, type of surgery, BMI before surgery and at start of HPN, indications for HPN including technical and nutritional complications (early within 2 months after surgery or late), outcome, PN regimen, and HPN complications. Patients were retrospectively included from January 2008 to June 2014. RESULTS: Eighteen HPN centers responded to the survey. A total of 2880 HPN patients were treated during the study period, 77 of whom had BS (65 females; mean age 51 ± 7 years); gastric bypass was performed in 69% of the patients; mean BMI was 44.4 before surgery and 23.2 at the start of HPN. Indications for HPN were early complications in 17 cases and late complications in 60 cases. Early complications were mostly anastomotic leakage/fistula; late complications were hypoalbuminemia, and vitamin and trace element deficiencies. Out of 77 patients, 16 needed a surgical re-intervention, 29 were weaned off HPN, and 6 died (no HPN-related deaths). During the HPN period, 58% of the patients were re-hospitalized and central venous complications were observed in 41%. Diabetes mellitus was described in 17/77 patients. HPN was supportive in 60 patients and exclusive in 17 patients (mean caloric intake: 23 ± 6 kcal/k BW/day and 1.2 g/kBW/day). Only 7/77 patients resumed their professional activities on HPN. CONCLUSIONS: This is the largest observational multicenter study describing the use of HPN in patients with post-BS complications. Severe hypoalbuminemia is a major late complication. Rates of re-hospitalization and CVC infection were high. HPN may be a "bridge therapy" before surgical revision after BS. The high mortality rate reflects the complexity of these cases.

A comparative study of peripherally-inserted and Broviac catheter complications in home parenteral nutrition patients.

BACKGROUND & AIMS: Peripherally inserted central venous catheters (PICC) have become increasingly popular for medium to long-term parenteral nutrition (PN) but there is limited data on the complication rates in this sub-group. We aimed to compare the rates of complications associated with tunneled catheters (Broviac) and PICC in home PN (HPN) patients. METHODS: All adult patients in an HPN program with a new Broviac or new PICC between 2009 and 2011 were included in this prospective observational study. Complication rates were compared by using Poisson regression and Kaplan Meier survival curves were used to compare the first complications that occurred. RESULTS: 204 catheters (133 Broviac and 71 PICC) were inserted in 196 adult patients. Mean follow-up from catheter insertions to their removal was 276 ± 219 days for Broviac (n = 86) vs. 74 ± 140.70 days for PICC (n = 56); p < 0.001. Complications were similar between Broviac and PICC (91/133 vs. 26/71). Catheter infection rate was lower in PICC (1.87 vs. 1.05 per 1000 catheter-days; p = 0.01). Catheter obstruction rates were similar for both catheters. Only PICC experienced venous thrombosis (0.4/1000). The proportion of catheters removed was lower in the Broviac group than in the PICC group (62.4% vs. 78.8%; p = 0.01) but those removed for complications were not different (28.6.7%vs. 25.3%; p = 0.64). CONCLUSIONS: In HPN patients, overall complications were similar in both the PICC and the Broviac groups. However, the Broviac catheter could be associated with an increase in catheter infection.

Home parenteral nutrition improves quality of life and nutritional status in patients with cancer: a French observational multicentre study.

PURPOSE: Malnutrition is a predictor of poor outcomes in patients with cancer. Little is known about the benefit of nutritional support in these patients. The purpose of this study was to assess the impact of home parenteral nutrition (HPN) on quality of life (Qol) in cancer patients. METHODS: We performed an observational prospective study to determine the impact of HPN on Qol in a population of patients with heterogeneous cancer. Physicians, patients and family members had to complete a questionnaire before HPN administration and 28 days after the course of HPN. Qol was evaluated using the self-administered questionnaire FACT-G. RESULTS: We included 767 patients with cancer of whom 437 ended the study. Mean patient age was 63±11.4 years and 60.5% were men. Primary gastrointestinal cancer was reported in 50% of patients and 65.3% were presenting metastases. Malnutrition was reported in 98.3%. After 28 days of HPN intake, significant improvement was observed in the Qol (49.95±5.82 vs. 48.35±5.01 at baseline, p<0.0001). The mean weight, serum albumin and the nutrition risk index had also improved significantly. Most patients (78%) had perceived a positive impact of the HPN. A significant improvement in patient's well-being was perceived also by family members and physicians. CONCLUSIONS: Our data suggest that preventing and correcting malnutrition using HPN in patients with cancer might have a significant benefit on their well-being. Randomized controlled studies are required to confirm this finding.

Impact of intravenous lipid emulsion infusion on the stratum corneum barrier function in patients receiving parenteral nutrition.

The importance of the lipid matrix of stratum corneum (SC) in epidermal barrier function is well documented. Intravenous lipid emulsions (ILE) provide essential fatty acids (EFAs), main components of the SC lipid matrix. The objective of this study was to investigate the influence of ILE upon SC barrier function. The skin barrier was assessed by measuring transepidermal water loss (TEWL). Patients receiving lipid-containing parenteral nutrition (LCPN) were compared to patients receiving lipid-free PN (LFPN). In addition, a before/after LCPN introduction study was set up to limit the influence of inter-individual variability. Twenty-six patients receiving LCPN and seven patients receiving LFPN were included. Median age was not significantly different between the two groups. The TEWL of the LCPN group (9.05 g/m2/h) was significantly lower than the TEWL of the LFPN group (12.1 g/m2/h; Wilcoxon test: p = 0.016). The relative variation of TEWL before and after ILE treatment of 5 studied patients was 21.29 ± 10.28 %. ILE improve epidermal barrier function when compared to lipid-free parenteral treatments. Results of the before/after study confirm this conclusion and the usefulness of ILE intake for preventing excessive TEWL. SC barrier function improvement could be a choice criterion between the different ILE generations, in particular in burn patients and premature neonates.

Is diabetes a risk factor for central venous access port-related bloodstream infection in oncological patients?

It was a dogma that patients with diabetes mellitus (DM) are at increased risk of infection or death associated with an infection. However, in cancer patients, this has not been well investigated. The aim was to investigate whether diabetic patients with cancer are at high risk of central venous access port (CVAP)-related bloodstream infection (BSI), and to analyse mortality after CVAP-BSI. A total of 17 patients with type 1 DM (T1DM), 66 with type 2 DM (T2DM) and 307 non-diabetic patients were included. Each patient was followed up until the first late CVAP-BSI or for a maximum for 1 year in the absence of a CVAP-BSI. Fifty-three CVAP-BSIs occurred in 66,528 catheter-days. The cumulative incidence of CVAP-BSI was not higher in T1DM (5.9 %; p = 0.17) and T2DM (19.7 %; p = 0.70) compared with the non-diabetic patients (12.7 %). However, in patients with CVAP-BSI, the 1-month crude mortality rate was higher in DM patients (42.9 % vs. 15.4 %; p = 0.04), whereas the mortality in patients without CVAP-BSI was similar in both groups of patients (19.8 % vs. 17.1 %; p = 0.58). Of the 12 deaths that occurred within 1 month of CVAP-BSI, 16.66 % was attributable to CVAP-BSI. The predictive factor of 1-month mortality was DM (p = 0.04). Parenteral nutrition (PN) was independently associated with CVAP-BSI in diabetic patients (p = 0.001). In this study, diabetes did not increase the risk of CVAP-BSI, but mortality was higher in diabetic patients who had a CVAP-BSI. This suggests, in addition to medical treatment, CVAP should be withdrawn after infection onset.

French clinical guidelines on perioperative nutrition. Update of the 1994 consensus conference on perioperative artificial nutrition for elective surgery in adults.

Surgical patient is a stressed patient. Aggression is more intense and prolonged as surgery is important. Surgery induces secretion of stress hormones, inflammatory mediators and metabolic changes resulting in significant catabolic phenomena. The presence of malnutrition is an independent risk factor for postoperative complications. Malnutrition increases morbidity (infections, delayed healing), mortality, length of stay and costs and impacts human quality of life for patients. It has been shown that the management of perioperative malnutrition reduces the additional risk generated by it. Perioperative nutritional support should not be systematically provided. Since 1994, recommendations on perioperative nutrition, the care of patients and the available resources have changed dramatically. An update of these recommendations was needed. In 2010, an expert panel of the French society of Anesthesiology (SFAR) and the French-speaking society of Clinical Nutrition and Metabolism (SFNEP) has made recommendations for good clinical practice of perioperative nutrition. They are presented. Thus, the perioperative nutritional management must be integrated in a process to reduce the operative risk: risk reduction due to preoperative malnutrition, reduced risk of postoperative malnutrition which may compromise the following treatments, reduction of postoperative metabolic complications, reducing the postoperative morbidity, especially infectious, through the use of pharmaconutrients either preoperatively or postoperatively in some patients.

[Nutritional assessment of patients before surgery]. / Évaluation nutritionnelle à la consultation d'anesthésie.

The identification of nutritional status is one of the objectives of the anaesthesia consultation often difficult to achieve routinely. It usually requires the use of multiple indicators, which are complex for a non-nutrition specialist. In preoperative period, nutritional assessment should be easy to do in order to identify patients who are malnourished or at risk of malnutrition and relevant information about nutritional risk should be registered in the patient chart. To facilitate this evaluation, we propose a stratification of nutritional risk in four grades (NG) using three types of simple and validated parameters: preoperative nutritional status (BMI, weight loss, eventually serum albumin), comorbidities and kind of surgery. This stratification can develop a tailored nutritional care for each patient.

Crohn's disease patients with chronic intestinal failure receiving long-term parenteral nutrition: a cross-national adult study.

BACKGROUND: Chronic intestinal failure (CIF) is a very rare Crohn's disease (CD) complication. AIM: To determine incidence of CIF treated with home parenteral nutrition (HPN) in adult CD patients and to isolate factors associated with severe CIF. METHODS: This retrospective multicentre study included 38 patients with CD-related CIF treated with HPN for at least 12 months in French HPN centres. Severe CIF was defined by a length of remnant small bowel of less than 100 cm or CIF occurrence within the 15 years following CD diagnosis. RESULTS: Median delay between CD diagnosis and CIF was 15 years. CIF incidence did not decrease over time (1.4/year before 1995 vs. 2.2/year after). Median number of small bowel resections per patient was three (range 1-8). Median small bowel resection, remnant and initial lengths were 160, 80 and 260 cm, respectively. Twenty-four per cent of patients developed stenosis within 1 year after CD diagnosis and 76% developed perforative complications within 2 years. In multivariate analysis, severe CIF, defined as CIF onset <15 years after CD diagnosis, was associated with a more recent CD diagnosis (odds ratio, 0.785; 95% confidence interval, 0.623-0.989). CIF occurred despite frequent use of immunosuppressants. Course of CD remained severe during HPN: immunosuppressants prescription occurred in 11 patients, surgery in six. Six patients died from CD (n = 2), HPN complications (n = 2) or other causes (n = 2). CONCLUSIONS: Chronic intestinal failure requiring HPN is rare during CD. Incidence remained stable over time. Surgical procedures play a minor role in the occurrence of severe chronic intestinal failure compared to CD severity.

[From the creation to the appreciation of a personal digital assistant-based clinical decision-support system for the management of artificial nutrition]. / De la conception à l'évaluation d'un programme d'aide à la prescription en nutrition artificielle sous Palm OS.

OBJECTIVES: The aim of our study was to assess the effect of NutriPDA, a personal digital assistant (PDA)-based clinical decision-support system (CDSS) for the management of artificial nutrition. A CDSS was developed and implemented on a handheld computer for use in the ICU after cardiovascular and thoracic surgery. STUDY DESIGN: System impact was assessed in a prospective "before/after" cohort trial. METHODS: After informed consent we studied 61 patients in the postcardiovascular and thoracic surgery ICU (age > 17 years, duration of artificial nutrition > 3 days, length of stay > 8 days). Patients were divided into two groups (before and after the use of NutriPDA: Group A: 32 patients (4-month period in 2005); group B: 29 patients (4-month period in 2006). RESULTS: There were no significant differences in anthropometric and clinical parameters between the 2 groups. Energetic intakes were < 80% of basal energetic expenditures in 21% and 1% of patients, respectively (P < 0.01). Caloric and nitrogen intakes were below international recommendation in Group A: 20+/-4 kcal/kg/d (mean+/-SD), 104+/-30 mg/kg/d, but not in Group B: 26+/-5 kcal/kg/d, 196+/-41 mg/kg/d (P < 0.01). CONCLUSION: NutriPDA was found to be able to optimize artificial nutrition by improving caloric intake in ICU. This new software has potential clinical applications.

[Impact of the guidelines on clinical practice of artificial nutrition in intensive care unit after cardiovascular and thoracic surgery]. / Impact d'un protocole de nutrition artificielle en réanimation après chirurgie cardiovasculaire et thoracique.

OBJECTIVE: To analyze the impact of an artificial nutrition program in post-anaesthesia intensive care unit. STUDY DESIGN: Observational study. PATIENTS AND METHODS: Patients with length of stay greater than 8 days after cardiovascular and thoracic surgery: Group 1: 34 patients (4-month period in 2000); group 2: 15 patients (2-month period in 2001); group 3: 40 patients (4-month period in 2003). Between these 3 periods, informations of physicians and written protocol in order to improve their nutritional knowledge. After analysis of variance (P<0.05). Newman-Keuls tests to compare themselves each groups. RESULTS: Anthropometric, demographic and clinical parameters were similar in the 3 groups. Energic intakes were less than 80% of basal energetic expenditures in 33%, 33 and 22% of patient, respectively (NS). Caloric and nitrogen intakes were less than recommended, respectively 19+/-6 (mean+/-SD), 21+/-7 and 21+/-8 kcal/kg/24 h and 102+/-32, 111+/-31 and 92+/-40 mg/kg/24 h (NS). However enteral nutrition was administered in 49, 40 and 100% of patients respectively (P<0.001). The glucid/lipid ratio improved from 0.47 in group 1 up to 0.68 in group 3 (P<0.0001). Vitamins, oligoelements and clinical and biological monitoring of artificial nutrition improved (P<0.001). CONCLUSION: A clinical audit demonstrated an improvement in artificial nutrition parameters but no significant change in others.

Structured triglyceride emulsions in parenteral nutrition.

Over the past 3 decades, various concepts for IV fat emulsions (IVFE) have been developed. A randomized, structured-lipid emulsion based on an old technology has recently become available. This structured-lipid emulsion is produced by mixing medium-chain triglycerides and long-chain triglycerides, then allowing hydrolysis to form free fatty acids, followed by random transesterification of the fatty acids into mixed triglyceride molecules. Studies in animals have shown an improvement in nitrogen balance with the use of these lipid emulsions. Only 8 human clinical studies with these products have been performed. The results of these human clinical studies have been less promising than the animal studies; however, an improvement in nitrogen balance and lipid metabolism exceeds results associated with infusion of long-chain triglycerides (LCT) or a physical mixture of long-chain triglycerides and medium-chain triglycerides (LCT-MCT). Structured-lipid emulsion seems to induce less elevation in serum liver function values compared with standard IVFEs. In addition, structured-lipid emulsions have no detrimental effect on the reticuloendothelial system. Further studies are necessary in order to recommend the use of structured-lipid emulsions. The clinical community hopes that chemically defined structured triglycerides will make it possible to determine the distribution of specific fatty acids on a specific position on the glycerol core and therefore obtain specific activity for a specific clinical situation.

Glucose-lipid ratio is a determinant of nitrogen balance during total parenteral nutrition in critically ill patients: a prospective, randomized, multicenter blind trial with an intention-to-treat analysis.

OBJECTIVE: Protein sparing, the major goal of nutritional support, may be affected by the glucose/lipid ratio. This study in critically ill patients compared the efficacy and tolerance of two isocaloric isonitrogenous total parenteral nutritions (TPN) having different glucose/lipid ratios. DESIGN: Multicentric prospective randomized study. PATIENTS: 47 patients with SAPS I score higher than 8 and requiring exclusive TPN. INTERVENTIONS: Patients received glucose/lipid ratios of 50/50 or 80/20. For 7 days all patients received 32 glucidolipidic kcal/kg and 0.27 g/kg nitrogen daily. All-in-one bags were prepared using industrial mixtures and a fat emulsion. MEASUREMENTS AND RESULTS: We determined TPN efficacy by nitrogen balance, urinary 3-methylhistidine/creatinine ratio, transthyretin and tolerance by glycemia, and liver enzymes. After controlling for five variables with significant effects, patients receiving the 50/50 ratio during TPN had significantly higher nitrogen balance than those receiving the 80/20 ratio. The daily difference in mean nitrogen sparing effect in favor of the latter group was 1.367 g (95% CI 0.0686-2.048). Glycemia on day 4 and gamma-glutamyltranspeptidase on day 8 were higher in group receiving the the 80/20 ratio. CONCLUSIONS: In critically ill patients TPN at a glucose/lipid ratio of 80/20 ratio induces a small nitrogen sparing effect compared to the ratio of 50/50, at the expense of poorer glycemic control. The clinical significance is unclear.

Specific changes in n -6 fatty acid metabolism in patients with chronic intestinal failure.

BACKGROUND AND AIMS: In patients presenting severe malabsorption, essential fatty acid (EFA) deficiency can be corrected by intravenous lipids, but EFA abnormalities persist. The purpose of this study was to evaluate the role of large resection of the small bowel or malabsorption on plasma phospholipid EFA profile. METHODS: The plasma phospholipid EFA composition was measured by gas chromatography in home parenteral nutrition patients with (n=13) or without small bowel resection (n=7) and in 14 healthy subjects. RESULTS: The two groups of patients had the same nutritional status and comparable amounts of intravenous fat. In both groups, plasma fatty acid concentrations were significantly different from those observed in healthy subjects without EFA deficiency. Among them: a decrease in 18:2n -6, 22:5n -3, 22:6n -3 and an increase in 18:3n -3, 20:4n -6, 22:4n -6. Moreover, arachidonic acid to linoleic acid ratio was higher in both groups of patients, suggesting a stimulation of the elongation and desaturation of 18:2n -6. In multiple linear regression, 18:2n -6 and 20:4n -6 levels were not associated with the small bowel length, only 22:6n -3 concentration was correlated with small bowel length. CONCLUSIONS: The patients with chronic intestinal failure on home parenteral nutrition presented specific change in their EFA and an increase in the n -6 fatty acid pathway. This could be related to the severe malabsorption.

The expression of the p85alpha subunit of phosphatidylinositol 3-kinase is induced by activation of the peroxisome proliferator-activated receptor gamma in human adipocytes.

AIMS/HYPOTHESIS: Thiazolidinediones are new oral antidiabetic drugs that activate the nuclear receptor PPARgamma. Our aim was to identify potential target genes of PPARgamma in the human adipocyte in order to clarify how thiazolidinediones improve insulin sensitivity. METHODS: The effect of BRL 49653 (Rosiglitazone) on the mRNA expression of insulin receptor, insulin receptor substrate-1, p85alpha, p110alpha and p110beta subunits of phosphatidylinositol 3-kinase, Glut 4 and hormone sensitive lipase was examined in isolated adipocytes. Target mRNA levels were determined by RT-competitive PCR. RESULTS: The BRL 49653 (1 micromol/l) increased the mRNA concentrations of p85alphaPI-3 K (264 +/- 46 vs 161 +/- 31 amol/microg total RNA, p = 0.003) whithout affecting the expression of the other mRNAs of interest. This effect was dose-dependent (K0.5 = 5 nmol/l) and was reproduced by a specific activator of RXR, indicating that it was probably mediated by the PPARgamma/RXR heterodimer. The BRL 49653 also increased the amount of p85alphaPI-3K protein in adipose tissue explants (71 +/- 19%). In addition, BRL 49653 produced a more than twofold increase in insulin stimulation of phosphatidylinositol 3-kinase activity and significantly enhanced the antilipolytic action of insulin. CONCLUSION/INTERPRETATION: This work demonstrates that the gene of p85alphaPI-3K is probably a target of PPARgamma and that thiazolidinediones can improve insulin action in normal human adipocytes. Although the precise mechanism of action of BRL 49653 on PI3-Kinase activity is not completely clear, these findings improve our understanding of the insulin-sensitizing effects of the thiazolidinediones, possible drugs for the treatment of Type II (non-insulin-dependent) diabetes mellitus.