Clinical approach to the management of Intestinal Failure Associated Liver Disease (IFALD) in adults: A position paper from the Home Artificial Nutrition and Chronic Intestinal Failure Special Interest Group of ESPEN.

We recommend that intestinal failure associated liver disease (IFALD) should be diagnosed by the presence of abnormal liver function tests and/or evidence of radiological and/or histological liver abnormalities occurring in an individual with IF, in the absence of another primary parenchymal liver pathology (e.g. viral or autoimmune hepatitis), other hepatotoxic factors (e.g. alcohol/medication) or biliary obstruction. The presence or absence of sepsis should be noted, along with the duration of PN administration. Abnormal liver histology is not mandatory for a diagnosis of IFALD and the decision to perform a liver biopsy should be made on a case-by-case basis, but should be particularly considered in those with a persistent abnormal conjugated bilirubin in the absence of intra or extra-hepatic cholestasis on radiological imaging and/or persistent or worsening hyperbilirubinaemia despite resolution of any underlying sepsis and/or any clinical or radiological features of chronic liver disease. Nutritional approaches aimed at minimising PN overfeeding and optimising oral/enteral nutrition should be instituted to prevent and/or manage IFALD. We further recommend that the lipid administered is limited to less than 1 g/kg/day, and the prescribed omega-6/omega-3 PUFA ratio is reduced wherever possible. For patients with any evidence of progressive hepatic fibrosis or overt liver failure, combined intestinal and liver transplantation should be considered.

Home Parenteral Nutrition in Adult Patients With Chronic Intestinal Failure: The Evolution Over 4 Decades in a Tertiary Referral Center.

BACKGROUND/AIMS: In Denmark, the public healthcare system ensures patients with intestinal failure (IF) the same rights for a life-saving treatment as patients with other organ failures. This study reports the epidemiological data from the largest Danish IF center. As one of the pioneering centers in treating IF with home parenteral nutrition (HPN), this study documents the HPN evolution and describes the demographics and outcome in one of the world's largest single-center cohorts. METHODS: We included patients with IF discharged with HPN from 1970-2010. Data were extracted according to European Society for Clinical Nutrition and Metabolism classifications from the Copenhagen IF database. RESULTS: Over the decades, we observed an exponential increase in the number of HPN patients. The 508 patients with IF collectively received HPN for 1751 years. While receiving HPN, 211 patients with IF (42%) died. Only 24 deaths were HPN related: sepsis (n = 10), liver disease (n = 12), central venous thrombosis (n = 1), and a complicated catheter placement (n = 1). The HPN-related mortality was as low as 0.014 deaths/HPN year. In the first decade, HPN was mainly provided to younger, intestinally resected adult patients with IF with inflammatory bowel disease (IBD), but numerically, they were subsequently outnumbered by elderly patients with IF with cancer or complications from non-IBD, noncancer abdominal surgery. Despite these demographic changes, the HPN-related mortality has decreased in the past decade. CONCLUSION: Evolving from being a rare, experimental treatment in the 1970s, HPN at present is safe with a low treatment-related mortality in the experienced center, despite HPN being more widely used in a more elderly population.

ESPEN guidelines on chronic intestinal failure in adults.

BACKGROUND & AIMS: Chronic Intestinal Failure (CIF) is the long-lasting reduction of gut function, below the minimum necessary for the absorption of macronutrients and/or water and electrolytes, such that intravenous supplementation is required to maintain health and/or growth. CIF is the rarest organ failure. Home parenteral nutrition (HPN) is the primary treatment for CIF. No guidelines (GLs) have been developed that address the global management of CIF. These GLs have been devised to generate comprehensive recommendations for safe and effective management of adult patients with CIF. METHODS: The GLs were developed by the Home Artificial Nutrition & Chronic Intestinal Failure Special Interest Group of ESPEN. The GRADE system was used for assigning strength of evidence. Recommendations were discussed, submitted to Delphi rounds, and accepted in an online survey of ESPEN members. RESULTS: The following topics were addressed: management of HPN; parenteral nutrition formulation; intestinal rehabilitation, medical therapies, and non-transplant surgery, for short bowel syndrome, chronic intestinal pseudo-obstruction, and radiation enteritis; intestinal transplantation; prevention/treatment of CVC-related infection, CVC-related occlusion/thrombosis; intestinal failure-associated liver disease, gallbladder sludge and stones, renal failure and metabolic bone disease. Literature search provided 623 full papers. Only 12% were controlled studies or meta-analyses. A total of 112 recommendations are given: grade of evidence, very low for 51%, low for 39%, moderate for 8%, and high for 2%; strength of recommendation: strong for 63%, weak for 37%. CONCLUSIONS: CIF management requires complex technologies, multidisciplinary and multiprofessional activity, and expertise to care for both the underlying gastrointestinal disease and to provide HPN support. The rarity of the condition impairs the development of RCTs. As a consequence, most of the recommendations have a low or very low grade of evidence. However, two-thirds of the recommendations are considered strong. Specialized management and organization underpin these recommendations.

Fifteen years' experience of intestinal and multivisceral transplantation in the Nordic countries.

OBJECTIVE: Intestinal and multivisceral transplantation have gained acceptance as treatment modalities for patients with: intestinal failure and life-threatening complications of parenteral nutrition (PN), rare cases of vascular abdominal catastrophes and selected cases of low-grade neoplastic tumors such as neuroendocrine pancreatic tumors and desmoids involving the mesenteric root. The aim was to describe the survival and nutritional outcome in the transplanted Nordic patients and the complications attributed to this procedure. METHOD: The authors included all Nordic patients transplanted between January 1998 and December 2013. Information on patients transplanted outside the Nordic region was collected through questionnaires. RESULTS: A total of 34 patients received different types of intestinal allografts. Currently, there are two Nordic transplant centers (n = 29) performing these procedures (Gothenburg, Sweden n = 24, Helsinki, Finland n = 5). The remaining five patients were transplanted in the USA (n = 3) and the UK (n = 2). Most patients were transplanted for life-threatening failure of PN (70%) caused primarily by intestinal motility diseases (59%). Allograft rejection was the most common complication and occurred in 79% of the patients followed by post-transplantation lymphoproliferative disorders (21%) and graft-versus-host disease (18%). The 1- and 5-year survival was 79% and 65% respectively for the whole cohort and nutritional autonomy was achieved in 73% of the adults and 57% of the children at 1 year after transplantation. CONCLUSION: This collective Nordic experience confirms that intestinal transplantation is a complex procedure with many complications, yet with the possibility to provide long-term survival in selected conditions previously considered untreatable.

A randomized, open-label, non-inferiority study of intravenous iron isomaltoside 1,000 (Monofer) compared with oral iron for treatment of anemia in IBD (PROCEED).

OBJECTIVES: In the largest head-to-head comparison between an oral and an intravenous (IV) iron compound in patients with inflammatory bowel disease (IBD) so far, we strived to determine whether IV iron isomaltoside 1,000 is non-inferior to oral iron sulfate in the treatment of iron deficiency anemia (IDA). METHODS: This prospective, randomized, comparative, open-label, non-inferiority study was conducted at 36 sites in Europe and India. Patients with known intolerance to oral iron were excluded. A total of 338 IBD patients in clinical remission or with mild disease, a hemoglobin (Hb) <12 g/dl, and a transferrin saturation (TSAT) <20% were randomized 2:1 to receive either IV iron isomaltoside 1,000 according to the Ganzoni formula (225 patients) or oral iron sulfate 200 mg daily (equivalent to 200 mg elemental iron; 113 patients). An interactive web response system method was used to randomize the eligible patient to the treatment groups. The primary end point was change in Hb from baseline to week 8. Iron isomaltoside 1,000 and iron sulfate was compared by a non-inferiority assessment with a margin of -0.5 g/dl. The secondary end points, which tested for superiority, included change in Hb from baseline to weeks 2 and 4, change in s-ferritin, and TSAT to week 8, number of patients who discontinued study because of lack of response or intolerance of investigational drugs, change in total quality of life (QoL) score to weeks 4 and 8, and safety. Exploratory analyses included a responder analysis (proportion of patients with an increase in Hb ≥2 g/dl after 8 weeks), the effect of regional differences and total iron dose level, and other potential predictors of the treatment response. RESULTS: Non-inferiority in change of Hb to week 8 could not be demonstrated. There was a trend for oral iron sulfate being more effective in increasing Hb than iron isomaltoside 1,000. The estimated treatment effect was -0.37 (95% confidence interval (CI): -0.80, 0.06) with P=0.09 in the full analysis set (N=327) and -0.45 (95% CI: -0.88, -0.03) with P=0.04 in the per protocol analysis set (N=299). In patients treated with IV iron isomaltoside 1,000, the mean change in s-ferritin concentration was higher with an estimated treatment effect of 48.7 (95% CI: 18.6, 78.8) with P=0.002, whereas the mean change in TSAT was lower with an estimated treatment effect of -4.4 (95% CI: -7.4, -1.4) with P=0.005, compared with patients treated with oral iron. No differences in changes of QoL were observed. The safety profile was similar between the groups. The proportion of responders with Hb ≥2 g/dl (IV group: 67%; oral group: 61%) were comparable between the groups (P=0.32). Iron isomaltoside 1,000 was more efficacious with higher cumulative doses of >1,000 mg IV. Significant predictors of Hb response to IV iron treatment were baseline Hb and C-reactive protein (CRP). CONCLUSIONS: We could not demonstrate non-inferiority of IV iron isomaltoside 1,000 compared with oral iron in this study. Based on the dose-response relationship observed with the IV iron compound, we suggest that the true iron demand of IV iron was underestimated by the Ganzoni formula in our study. Alternative calculations including Hb and CRP should be explored to gauge iron stores in patients with IBD.

State of the iron: how to diagnose and efficiently treat iron deficiency anemia in inflammatory bowel disease.

Iron deficiency anemia (IDA) frequently occurs in patients suffering from inflammatory bowel disease (IBD) and negatively impacts their quality of life. Nevertheless, the condition appears to be both under-diagnosed and undertreated. Regular biochemical screening of patients with IBD for anemia by the gastroenterology community has to be advocated. Oral iron is a low cost treatment however its effectiveness is limited by low bioavailability and poor tolerability. Intravenous (IV) iron rapidly replenishes iron stores and has demonstrated its safe use in a number of studies in various therapeutic areas. A broad spectrum of new IV iron formulations is now becoming available offering improved tolerability and patient convenience by rapidly restoring the depleted iron status of patients with IBD. The following article aims to review the magnitude of the problem of IDA in IBD, suggest screening standards and highlight existing and future therapies.

Four-week parenteral nutrition using a third generation lipid emulsion (SMOFlipid)--a double-blind, randomised, multicentre study in adults.

PRECIS: The aim of this study was to evaluate the safety and tolerance of a soybean/MCT/olive/fish oil emulsion in intestinal failure patients on long-term parenteral nutrition. 73 patients took part in a randomized, double-blind, multi-centre study. The study demonstrates that the lipid emulsion containing four different types of oils is safe and well tolerated in long-term PN. BACKGROUND & AIM: Long-term safety and efficacy of a lipid emulsion containing soybean oil, medium-chain triglycerides (MCT), olive oil and fish oil and enriched in vitamin E have not yet been evaluated in adult patients requiring long-term parenteral nutrition (PN). METHODS: Randomised, controlled, double-blind, multicentre study in 73 patients with stable intestinal failure, requiring PN with either soybean/MCT/olive/fish emulsion (SMOFlipid, n = 34) or soybean emulsion (Intralipid, control n = 39) for 4 weeks. Safety and tolerance were monitored with standard clinical laboratory parameters, adverse events (AEs, according to the Common Terminology Criteria for Adverse Events (CTCAE) classification v 3.0) and vital signs. Fatty acid pattern in red blood cell phospholipids and plasma lipoproteins, serum Vitamin E, Interleukin (IL)-6, and soluble tumour necrosis (s-TNF)-receptor(R)II were also evaluated. RESULTS: Mean concentrations of alanine transaminase (ALT), aspartate transaminase (AST) and total bilirubin, whilst remaining within the reference range, were significantly lower with soybean/MCT/olive/fish (SMOF) oil emulsion after the treatment period compared to control. Eicosapentaenoic acid, docosahexaenoic acid and n-3/n-6 fatty acid ratio increased in the SMOF group, while they remained unchanged in the control in plasma and RBC. Serum α-tocopherol concentrations significantly increased in the study group compared to control (p = 0.0004). IL-6 and sTNF-RII levels did not change during the study period. Grade 4 (serious) adverse events occurred in 2 SMOF patients and in 8 control patients (p = 0.03). CONCLUSIONS: Soybean/MCT/olive/fish emulsion was safe and well tolerated over 4 weeks and leads to positive change in fatty acids profile.

Diagnosis and treatment of fistulising Crohn's disease.

A fistula is defined as a pathological connection between the intestine and an inner (bladder or other intestine) or outer (vagina or skin) epithelial surface. Fistulas are discovered in up to 25% of all Crohn's disease patients during long-term follow-up examinations. Most are perianal fistulas, and these may be classified as simple or complex. The initial investigation of perianal fistulas includes imaging (MRI of the pelvis and rectum), examination under anaesthesia (EUA) with digital imaging, endoscopy, probing and anal ultrasound. Non-perianal fistulas require contrast imaging and/or CT/MRI for complete anatomical definition. Any abscess should be drained, and the disease extent throughout the entire gastrointestinal tract should be evaluated. Treatment goals for perianal fistulas include reduced fistula secretion or none, evaluated by clinical examination; the absence of abscesses; and patient satisfaction. MR imaging is required to demonstrate definitive fistula closure. Fistulotomy is considered for simple perianal fistulas. In complex perianal fistulas, antibiotics and azathioprine or 6-mercaptopurine, which are often combined with a loose seton, constitute the first-line medical therapy. In cases with persistent secretion, infliximab at 5 mg/kg is given at weeks 0, 2, and 6 and subsequently every 8 weeks. Adalimumab may improve fistula response in both infliximab-naïve patients and following infliximab treatment failure. Local therapy with fibrin glue or fistula plugs is rarely effective. Definitive surgical closure of perianal fistulas using an advancement flap may be attempted, but this procedure is associated with a high risk of relapse. Colostomy and proctectomy are the ultimate surgical treatment options for fistulas. Intestinal resection is almost always needed for the closure of symptomatic non-perianal fistulas.

Management and perception of hospital undernutrition-a positive change among Danish doctors and nurses.

BACKGROUND: Undernutrition in hospitals is a common problem associated with increased morbidity and mortality, prolonged convalescence and duration of hospital stay and increased health care costs. During recent years several initiatives have brought hospital undernutrition into focus and guidelines and standards have been published. In 1997, a questionnaire-based survey among Danish hospital doctors and nurses in selected departments concluded that clinical nutrition did not fulfil accepted standards. AIMS: We wished to determine if improvements had occurred in the intervening period. METHOD: Thus, in 2004 a similar questionnaire was sent to 4000 randomly selected Danish hospital doctors and nurses and responses were compared to those from 1997. The questionnaire dealt with attitudes and practice in the areas of nutritional screening, treatment plan, monitoring as well as with knowledge, education, tools and guidelines, organisation and possible barriers to implementation of nutritional screening and therapy. RESULTS: The overall response rate was 38%. We observed a marked improvement especially in screening procedures, calculation of energy intake in at-risk patients and local availability of guidelines. Many departments had appointed staff members with special interest and knowledge in clinical nutrition. CONCLUSION: Although significant positive changes had thus occurred, the main barriers against implementation of good nutrition care continued to be lack of knowledge, interest and responsibility, in combination with difficulties in making a nutrition plan. This will be the focus of future activities.

Monitoring of patients on home parenteral nutrition (HPN) in Europe: a questionnaire based study on monitoring practice in 42 centres.

INTRODUCTION & AIM: To gather information about how adult patients on home parenteral nutrition (HPN) are monitored after discharge from the hospital. METHOD: A questionnaire about HPN monitoring practice was circulated to HPN centres in eight European countries through the representative of the ESPEN HAN-working group. Centres were asked about guidelines, home visits and how monitoring and handling of complications were managed. RESULTS: Fourty-two centres in eight European countries completed the questionnaire: UK n=14, France n=9, Belgium n=4, Italy n=4, Poland n=4, Denmark n=4, Spain n=2, Germany n=1. The HPN experience of the centres was in the range 2-30 years. Centres ranged in size from 0 to 125 HPN patients representing a total number of 934 of whom 54% had received HPN for more than 2 years. The primary disease was non-malignant in 90% whilst 10% had been diagnosed with active cancer. Of the centres 92% had a HPN team and 66% had written guidelines for monitoring HPN. Home visits after discharge for monitoring purposes were carried out by 31 of the centres involving the HPN team, general practitioner, community nurse or home care agency. Stable patients on HPN for more than 12 months were monitored at the discharging hospital (73%), at a local hospital (12%), by the General Practitioner (11%) or by a home care agency (4%). Of the centres, 90% reported that the main responsibility for monitoring was assigned to a specific person. The intervals between monitoring visits for the stable HPN patient was in the range 1-6 months, 52% of the centres reported intervals of 2-3 months. In case of complications 76% of centres reported that patients got in touch with the HPN team, 2% the local hospital, 5% the home care agency, and 17% other. Re-admission to hospital was usually to the HPN centre and only occasionally to a local hospital. CONCLUSION: In Europe a specialised team at the discharging hospital monitors HPN patients and 66% of the centres had some kind of written guidelines.

Home parenteral nutrition in Denmark in the period from 1996 to 2001.

OBJECTIVE: Home parenteral nutrition (HPN) has been provided to patients with intestinal failure in Denmark since 1970. The results of a national survey comprising a well-defined cohort receiving treatment with HPN in Denmark in the period from 1996 to the end of 2000 are presented, including data on incidence, prevalence, patient characteristics and complications of HPN treatment. MATERIAL AND METHODS: HPN was given to 202 patients (115 F, 87 M) 34% with short-bowel syndrome due to inflammatory bowel disease, 26% with cancer, 22% with surgical complications and 19% with other causes, for a total of 410 catheter years. RESULTS: At the end of 2000, the prevalence of HPN in Denmark was 19.2 per million, and the average annual incidence was 5.0 per million per year over the 5-year period. The incidence rates of catheter-related bacteraemia and thrombosis were 0.48 and 0.02 episodes per catheter year, respectively. The average catheter lasted 1.5 years. Venous inaccessibility was never encountered. The 5-year mortality was 25% (n = 51). However, only six deaths were related to HPN. Three patients died of HPN-related liver failure, two died of catheter-related sepsis and one patient died of an embolus induced by a catheter thrombus. CONCLUSIONS: Since its introduction in 1970 in Denmark, HPN has evolved from an experimental approach to a standardized therapy for patients with intestinal failure. Although HPN morbidity is increasing slightly compared with that reported in earlier surveys, HPN-induced mortality is still less than 1% per year.

A randomized, placebo-controlled trial of certolizumab pegol (CDP870) for treatment of Crohn's disease.

BACKGROUND & AIMS: To investigate the efficacy and safety of certolizumab pegol (a polyethylene-glycolated Fab' fragment of anti-tumor necrosis factor, CDP870) in Crohn's disease. METHODS: In a placebo-controlled, phase II study, 292 patients with moderate to severe Crohn's disease received subcutaneous certolizumab 100, 200, or 400 mg or placebo at weeks 0, 4, and 8. The primary end point was the percentage of patients with a clinical response at week 12 (a Crohn's Disease Activity Index decrease of > or = 100 points or remission [Crohn's Disease Activity Index < or = 150 points]) in the intent-to-treat population. RESULTS: All certolizumab doses produced significant clinical benefit over placebo at week 2 (placebo, 15.1%; certolizumab 100 mg, 29.7% [P = .033]; 200 mg, 30.6% [P = .026]; 400 mg, 33.3% [P = .010]). At all time points, the clinical response rates were highest for certolizumab 400 mg, greatest at week 10 (certolizumab 400 mg, 52.8%; placebo, 30.1%; P = .006) but not significant at week 12 (certolizumab 400 mg, 44.4%; placebo, 35.6%; P = .278). Patients with baseline C-reactive protein levels of 10 mg/L or greater (n = 119) showed clearer separation between active treatment and placebo (week 12 clinical response: certolizumab 400 mg, 53.1%; placebo, 17.9%; P = .005; post hoc analysis) owing to a lower placebo response rate than patients with C-reactive protein levels of less than 10 mg/L. Adverse events were similar among groups. CONCLUSIONS: Certolizumab 400 mg may be effective and is well tolerated in patients with active Crohn's disease. High placebo response rates in the large patient subgroup with low C-reactive protein levels may have obscured statistical separation between certolizumab and placebo. Ongoing phase III trials are necessary to establish the clinical efficacy of certolizumab.

Assessment of the longitudinal changes in bone mineral density in patients receiving home parenteral nutrition.

BACKGROUND: Low bone mineral density (BMD) is commonly reported in patients receiving home parenteral nutrition (HPN), but it remains unclear whether or not an accelerated bone loss occurs during HPN therapy. We evaluated the spinal, hip, and forearm bone mass density longitudinally in a cohort of 75 patients receiving HPN. METHODS: A total of 943 regional dual-energy x-ray absorptiometry scans, 335 spinal, 318 hip, and 290 forearm, obtained between 1995 and 2003 in 75 patients receiving HPN, were used for the analysis of the annual changes in BMD. The average (SD) number of scans per patients was 4.4 (2.9), and follow-up time was 4.1 (1.9) years. Diagnoses were Crohn's disease (n = 35) and other conditions (non-Crohn's diseases; n = 40). Data were analyzed using a linear random coefficient model. RESULTS: There was a statistically significant overall decline over time in spinal, hip, and forearm BMD, corresponding roughly to a 1% annual loss (p < .005); however, the loss was not significantly larger than that of age and sex-matched healthy subjects. In Crohn's disease patients, model estimates of spinal and hip BMD on the initiation of HPN therapy were significantly reduced compared with normal, whereas values were not significantly reduced in non-Crohn's disease patients. CONCLUSIONS: With the current protocols for HPN treatment, the annual decline in BMD is moderate and not significantly larger than in age- and sex-matched healthy subjects. A considerable part of the metabolic bone disease in these patients is related to the underlying disease for which the HPN was indicated.

Prevalence of patients at nutritional risk in Danish hospitals.

BACKGROUND & AIMS: Undernutrition is associated with increased morbidity and mortality and is common in patients admitted to hospital. We examined (1) the prevalence of patients at nutritional risk, (2) whether these patients were identified by the staff, and (3) whether a nutritional plan and monitoring was made for patients at nutritional risk. METHODS: A cross-sectional study in 15 randomly selected departments (>200 beds, departments of internal medicine, gastro- and orthopedic surgery) in Danish hospitals. The patients were characterized by scoring the components 'undernutrition' and 'severity of disease' in 4 categories (absent, mild, moderate or severe). The patient could have a score of 0-3 for each component (undernutrition and severity of disease), and any patient with a total score > or = 3 was considered at nutritional risk. Undernutrition was evaluated by 3 variables (BMI, recent weight loss, recent food intake). RESULTS: Out of 590 patients, 39.9% were nutritionally at risk, with the highest prevalence in departments of gastro-surgery (57%). BMI was <18.5 in 10.9%, and between 18.5 and 20.5 in 16.7% of the patients. In 7.6% the records contained information about nutritional risk, in 14.2% about a nutrition plan of which only 55.2% included a plan for monitoring. Measurements of BMI were found in 3% of the records. Both severity of disease (P < 0.02) and weightloss (P < 0.04) were predictive for making a nutrition plan. CONCLUSIONS: Nearly 40% of patients in departments of internal medicine, gastro- and orthopedic surgery are at nutritional risk, and only a minor part of these patients are identified. As a consequence only few patients at a nutritional risk have a nutrition plan and a plan for monitoring.

[Pregnancy-related thrombocytosis]. / Svangerskabsrelateret trombocytose.

Pregnancy-related thrombocythaemia comprises myeloproliferative and inflammatory reactive subsets. In pregnant women treated for myeloproliferative disorders, especially polycythaemia vera and primary thrombocytosis, only 50-70 per cent are delivered successfully of a normal healthy baby. The maternal complications are cerebral, cardiac, and abdominal arterial thrombosis, and with deep venous thrombosis of the legs, whereas bleedings are mainly seen in the case of extreme thrombocythaemia, owing to absorption of factors by the platelets. The foetal complication are dominated by abruptio placentae, pre-eclampsia, placental insufficiency, and death. Reactive thrombocythaemia includes the physiological rise in platelets postpartum, believed to be part of the normal maternal haemostasis, which almost never causes thromboembolic complications, as far as is known today. In contrast, the inflammatory reactive thrombocythaemia, related to severe foetal and/or maternal necrosis, is generally related only to a moderate rise in the platelet count. As the blood-platelet count does not appear to be routine at general pregnancy check-ups, it is necessary to be aware of risk groups, consisting of women with otherwise unexplained abortions or stillbirths, unexplained foetal and placental malformations, and pre-eclampsia, even if the woman has never had any thromboembolic complications.