1. Introduction Neonatal sepsis is a clinical syndrome of bacteremia with systemic symptoms and signs of infection in the first 28 days of life. In recent national studies, the incidence of suspected neonatal sepsis among admitted neonates varied from 32.9% to 45.9% with a higher incidence in preterm.

Several soluble adhesion molecules are released in neonatal sepsis. Among these, soluble intercellular adhesion molecule 1 (sICAM-1) and leukocyte integrin β2 are early predictors of sepsis with high sensitivity and specificity.

Septic preterm infants are more vulnerable to undernutrition and postnatal failure of growth requiring more nutritional support. The early use of adequate parenteral nutrition (PN) minimizes weight loss, improves growth and neurodevelopmental outcomes, and appears to reduce the risk of mortality. In preterm neonates, lipid emulsions (LE) play an important role in PN being an important source of energy, fat soluble vitamins and essential fatty acids. For the last few decades, pure soyabean oil based LE (S-LE) were used worldwide and consisted of long chain fatty acids with omega 3: omega 6 ratio of 1:5.5. Recently, LE preparations derived from multiple sources have been developed for clinical application. MOFS-LE mixture is one containing a mixture of 30% medium chain triglycerides (MCT), 25% olive oil, 15% fish oil, and 30% soyabean oil. They are supposed to have better immunomodulatory and anti-inflammatory properties with fewer side effects.

There are several studies concerning the comparison between of S-LE and MOFS-LE as regards the efficacy and safety on neonates but without a consensus on the ideal LE .To our knowledge, this is the first randomized controlled trial (RCT) of two different intravenous LE conducted in septic preterm infants to unravel the short term effects of S-LE versus MOFS-LE on the outcomes of neonatal sepsis as well as on the serum levels of sICAM-1 and leukocyte integrin β2.

2. Materials and methods

2.1 Study design and participants Our study is a randomized controlled pilot trial that is conducted in the Neonatal Care Unit (NCU) of Mansoura University Children's Hospital, Egypt including 40 preterm infants. The study was accepted by International research board of Medical Faculty of Mansoura University.

Sepsis is defined clinically as the presence of three or more of the following groups of clinical signs: tachypnea >> 60 breath/min), retractions, nasal flaring, apnea, cyanosis; bradycardia (<100 beat/min), tachycardia >> 180 beat/min); seizures, hypotonia; poor skin color, capillary refill time>> two seconds; lethargy, irritability.

2.2 Randomization Informed consents are obtained from the parents. Then, patients are assigned blindly into one of two groups using cards provided in opaque sealed consecutively numbered envelopes. MOFS group (n=20 cases) receive PN containing MOFS-LE which is a 30:30:25:15 mixture of soybean oil, MCT, olive oil, and fish oil (SMOFlipid ®, Fresenius kabi, Uppsala, Sweden) for seven consecutive days. S group (n=20 cases) receive PN containing S-LE (20% Intralipid ®, Fresenius kabi, Uppsala, Sweden) for seven consecutive days.

PN preparation is carried out by a special nurse under complete aseptic technique in a separate room and provided as (all in one) admixture. The all in one admixture is prepared daily following usual unit policy except for the type of intravenous LE and administrated through central venous catheters at a constant (pump controlled) rate for 24 hours per day. PN administration to preterm infants will be given according to the standard protocol of the unit. The starting dose of intravenous lipid is 0.5 g/kg/day on the first day of PN, increase gradually by 1gm/kg/day to a maximum dose of 3.5 gm/kg/day as recommended by the European Society for Clinical Nutrition and Metabolism (ESPEN)/European Society of Paediatric Gastroenterology, Hepatology, and Nutrition. As a rule, when serum triglyceride levels exceed 250 mg/dl, the lipid dosage will be reduced by 25% of the given dose. Amino acids, carbohydrates, trace elements, vitamins and electrolytes are given for both groups according to the unit's standard protocol.

Empirical antibiotics are started for the cases at the onset of clinically suspected sepsis according to our NCU policy. Cases of EOS receive ampicillin and gentamycin while cases of LOS receive with either cefotaxime or gentamycin.

2.3 Interventions Epidemiological and clinical data are collected including gestational age (assessed by early ultrasound scan, menstrual history and the new Ballard score), postnatal age, sex, birth weight, maternal age, type of delivery, risk factors for sepsis (maternal fever, urinary tract infection, prolonged rupture of membranes, central line insertion or surgical operation), duration of antibiotic treatment, mechanical ventilation, duration of hospital stay and mortality rate. Clinical examination is done on 1st day of suspicion of sepsis and seven days later including growth parameters (body weight, length and head circumference), vital signs, cardiovascular examination (skin perfusion and pulsations), chest examination (signs of respiratory distress and abnormal adventitious sounds), gastrointestinal system (abdominal distention, signs of feeding intolerance and organomegaly) as well as central nervous system examination (activity, neonatal reflexes, abnormal tone and seizures).

Laboratory investigations is done on 1st day of suspicion of sepsis and seven days post-randomization. Five ml of venous blood was withdrawn, 1 ml of them into Ethylene diamine tetra acetic acid (EDTA) tube for complete blood count (CBC), 1 ml blood into BACTEC vial for blood culture and 3 ml blood into a plain tube which was centrifuged and the resulting serum will be used for assay of C-reactive protein (CRP), random blood glucose, serum creatinine and serum triglyceride, sICAM-1 and leukocyte integrin ß2. Urine culture and cerebrospinal fluid examination will be done for LOS. Serum levels of sICAM-1 and leukocyte integrin ß2 are quantitatively measured by enzyme-linked immunosorbent assay (ELISA) technique.

2.4 Statistical analysis Statistical Package for the Social Sciences (SPSS) version 22 (IBM Corporation,Newyork, USA) was used to analyze data. Frequency (number and percent) will be used to express categorical variables. Chi square test is used for comparison of categorical variables between two groups. Fisher exact test is needed when more than 25% of the cells have expected count less than 5. Non-parametric data are presented as median and range, while parametric data are presented as mean ± standard deviation (SD). Between-groups comparisons are done using Student's t test (for parametric data) and Mann-Whitney test (for non-parametric data). Within-group comparisons of changes from 1st day to 7th day are carried out by means of Wilcoxon test for non-parametric data and paired t-test for parametric data.