Preterm Formula, Fortified or Unfortified Human Milk for Very Preterm Infants, the PREMFOOD Study: A Parallel Randomised Feasibility Trial.

OBJECTIVE: Uncertainty exists regarding optimal supplemental diet for very preterm infants if the mother's own milk (MM) is insufficient. We evaluated feasibility for a randomised controlled trial (RCT) powered to detect important differences in health outcomes. METHODS: In this open, parallel, feasibility trial, we randomised infants 25+0-31+6 weeks of gestation by opt-out consent to one of three diets: unfortified human milk (UHM) (unfortified MM and/or unfortified pasteurised human donor milk (DM) supplement), fortified human milk (FHM) (fortified MM and/or fortified DM supplement), and unfortified MM and/or preterm formula (PTF) supplement from birth to 35+0 weeks post menstrual age. Feasibility outcomes included opt-outs, adherence rates, and slow growth safety criteria. We also obtained anthropometry, and magnetic resonance imaging body composition data at term and term plus 6 weeks (opt-in consent). RESULTS: Of 35 infants randomised to UHM, 34 to FHM, and 34 to PTF groups, 21, 19, and 24 infants completed imaging at term, respectively. Study entry opt-out rate was 38%; 6% of parents subsequently withdrew from feeding intervention. Two infants met predefined slow weight gain thresholds. There were no significant between-group differences in term total adipose tissue volume (mean [SD]: UHM: 0.870 L [0.35 L]; FHM: 0.889 L [0.31 L]; PTF: 0.809 L [0.25 L], p = 0.66), nor in any other body composition measure or anthropometry at either timepoint. CONCLUSIONS: Randomisation to UHM, FHM, and PTF diets by opt-out consent was acceptable to parents and clinical teams, associated with safe growth profiles and no significant differences in body composition. Our data provide justification to proceed to a larger RCT.

Accuracy of collagen fibre estimation under noise using directional MR imaging.

In tissues containing significant amounts of organised collagen, such as tendons, ligaments, menisci and articular cartilage, MR imaging exhibits a strong signal intensity variation caused by the angle between the collagen fibres and the magnetic field. By obtaining scans at different field orientations it is possible to determine the unknown fibre orientations and to deduce the underlying tissue microstructure. Our previous work demonstrated how this method can detect ligament injuries and maturity-related changes in collagen fibre structures. Practical application in human diagnostics will demand minimisation of scanning time and likely use of open low-field scanners that can allow re-orienting of the main field. This paper analyses the performance of collage fibre estimation for various image SNR values, and in relation to key parameters including number of scanning directions and parameters of the reconstruction algorithm. The analysis involved Monte Carlo simulation studies which provided benchmark performance measures, and studies using MR images of caprine knee samples with increasing levels of synthetic added noise. Tractography plots in the form of streamlines were performed, and an Alignment Index (AI) was employed as a measure of the detected orientation distribution. The results are highly encouraging, showing high accuracy and robustness even for low image SNR values.

Detection of maturity and ligament injury using magic angle directional imaging.

PURPOSE: To investigate whether magnetic field-related anisotropies of collagen may be correlated with postmortem findings in animal models. METHODS: Optimized scan planning and new MRI data-processing methods were proposed and analyzed using Monte Carlo simulations. Six caprine and 10 canine knees were scanned at various orientations to the main magnetic field. Image intensities in segmented voxels were used to compute the orientation vectors of the collagen fibers. Vector field and tractography plots were computed. The Alignment Index was defined as a measure of orientation distribution. The knees were subsequently assessed by a specialist orthopedic veterinarian, who gave a pathological diagnosis after having dissected and photographed the joints. RESULTS: Using 50% less scans than reported previously can lead to robust calculation of fiber orientations in the presence of noise, with much higher accuracy. The 6 caprine knees were found to range from very immature (< 3 months) to very mature (> 3 years). Mature specimens exhibited significantly more aligned collagen fibers in their patella tendons compared with the immature ones. In 2 of the 10 canine knees scanned, partial cranial caudal ligament tears were identified from MRI and subsequently confirmed with encouragingly high consistency of tractography, Alignment Index, and dissection results. CONCLUSION: This method can be used to detect injury such as partial ligament tears, and to visualize maturity-related changes in the collagen structure of tendons. It can provide the basis for new, noninvasive diagnostic tools in combination with new scanner configurations that allow less-restricted field orientations.

Development of Early Adiposity in Infants of Mothers With Gestational Diabetes Mellitus.

OBJECTIVE: Infants born to mothers with gestational diabetes mellitus (GDM) are at greater risk of later adverse metabolic health. We examined plausible candidate mediators, adipose tissue (AT) quantity and distribution and intrahepatocellular lipid (IHCL) content, comparing infants of mothers with GDM and without GDM (control group) over the first 3 postnatal months. RESEARCH DESIGN AND METHODS: We conducted a prospective longitudinal study using MRI and spectroscopy to quantify whole-body and regional AT volumes, and IHCL content, within 2 weeks and 8-12 weeks after birth. We adjusted for infant size and sex and maternal prepregnancy BMI. Values are reported as the mean difference (95% CI). RESULTS: We recruited 86 infants (GDM group 42 infants; control group 44 infants). Mothers with GDM had good pregnancy glycemic control. Infants were predominantly breast-fed up to the time of the second assessment (GDM group 71%; control group 74%). Total AT volumes were similar in the GDM group compared with the control group at a median age of 11 days (-28 cm(3) [95% CI -121, 65], P = 0.55), but were greater in the GDM group at a median age of 10 weeks (247 cm(3) [56, 439], P = 0.01). After adjustment for size, the GDM group had significantly greater total AT volume at 10 weeks than control group infants (16.0% [6.0, 27.1], P = 0.002). AT distribution and IHCL content were not significantly different at either time point. CONCLUSIONS: Adiposity in GDM infants is amplified in early infancy, despite good maternal glycemic control and predominant breast-feeding, suggesting a potential causal pathway to later adverse metabolic health. Reduction in postnatal adiposity may be a therapeutic target to reduce later health risks.

Avoiding sedation in research MRI and spectroscopy in infants: our approach, success rate and prevalence of incidental findings.

Performing magnetic resonance investigations in a paediatric population can be difficult; image acquisition is commonly complicated by movement artefact and non-compliance. Sedation is widely used for clinically indicated investigations, but there is controversy when used for research imaging. Over a 10-year period we have performed whole body MRI on over 450 infants and hepatic magnetic resonance spectroscopy on over 270 infants. These investigations have been accomplished without the use of sedation in infants up to 3 months of age. Our overall success rate in achieving good quality images free of movement artefact is 94%. The prevalence of incidental findings on whole body (excluding brain) MRI in our cohort was 0.8%. We conclude that the use of sedation for research MRI in this group is not necessary. Our approach to MRI in infancy is also described.

A method for measuring the cross sectional area of the anterior portion of the optic nerve in vivo using a fast 3D MRI sequence.

PURPOSE: To investigate the three-dimensional (3D) fast-recovery fast spin-echo accelerated (FRFSE-XL) sequence as a new application for measuring the intraorbital optic nerve (ION) mean cross-sectional area in vivo and to determine its value within a commonly used high resolution imaging protocol. MATERIALS AND METHODS: The entire ION was scanned in nine healthy volunteers (mean age 32 +/- 4 years) using the 3D FRFSE-XL sequence and a commonly used high resolution short-echo fast fluid-attenuated inversion-recovery (sTE fFLAIR) sequence with identical slice locations at 1.5T. The mean cross-sectional area from both sequences was measured on a slice-by-slice basis from 3 mm behind the globe to the orbital apex. The reproducibility of both techniques was assessed from repeated scans (scan-rescan) and repeated image analysis (intraobserver). RESULTS: Measurement of the mean cross-sectional area of the anterior 9 mm segment of the ION was only possible using the 3D FRFSE-XL sequence with a mean area of 11.6 +/- 2.2 mm(2) (scan rescan COV = 3.3 +/- 1.5, intraobserver COV = 2.4 +/- 0.02) whereas the remainder segment of the ION (i.e., 9 mm behind the globe to the orbital apex) could only be measured with the use of the sTE fFLAIR with a mean area of 8.5 +/- 1.7 mm(2) (scan rescan COV = 4.9 +/- 2.5 and intraobserver COV = 3.70 +/- 0.03). CONCLUSION: The 3D FRFSE-XL allows fast and reproducible measurement of the cross-sectional area of the anterior 9 mm segment of the ION, which is not possible using commonly used imaging sequences due to image degradation from motion, and is of complementary value to the existing imaging protocol for ION atrophy quantification.

Magnetic resonance imaging of cortical bone with ultrashort TE pulse sequences.

PURPOSE: Normal adult cortical bone has a very short T(2) and characteristically produces no signal with pulse sequence echo times (TEs) routinely used in clinical practice. We wished to determine whether it was possible to use ultrashort TE (UTE) pulse sequences to detect signal from cortical bone in human subjects and use this signal to characterise this tissue. SUBJECTS AND METHODS: Seven volunteers and 10 patients were examined using ultrashort TE pulse sequences (TE=0.07 or 0.08 ms). Short and long inversion as well as fat suppression pulses were used as preparation pulses. Later echo images were also obtained as well as difference images produced by subtracting a later echo image from a first echo image. Saturation pulses were used for T(1) measurement and sequences with progressively increasing TEs for T(2)* measurement. Intravenous gadodiamide was administered to four subjects. RESULTS: Signal in cortical bone was detected with UTE sequences in children, normal adults and patients. This signal was usually made more obvious by subtracting a later echo image from the first provided that the signal-to-noise ratio was sufficiently high. Normal mean adult T(1)s ranged from 140 to 260 ms, and mean T(2)*s ranged from 0.42 to 0.50 ms. T(1) increased significantly with age (P<.01). Increased signal was observed after contrast enhancement in the normal volunteer and the three patients to whom it was administered. Reduction in signal from short T(2) components was seen in acute fractures, and increase in signal in these components was seen with new bone formation after fracture malunion. In a case of osteoporosis, bone cross-sectional area and signal level appeared reduced. CONCLUSION: Signal can be detected from normal and abnormal cortical bone with UTE pulse sequences, and this can be used to measure its T(1) and T(2)* as well as observe contrast enhancement. Difference images are of value in increasing the conspicuity of cortical bone and observing abnormalities in disease.

Magic angle effects in MR neurography.

BACKGROUND AND PURPOSE: Magic angle effects are well recognized in MR imaging of tendons and ligaments, but have received virtually no attention in MR neurography. We investigated the hypothesis that signal intensity from peripheral nerves is increased when the nerve's orientation to the constant magnetic induction field (B(0)) approaches 55 degrees (the magic angle). METHODS: Ten volunteers were examined with their peripheral nerves at different orientations to B(0) to detect any changes in signal intensity and provide data to estimate T2. Two patients with rheumatoid arthritis also had their median nerves examined at 0 degrees and 55 degrees. RESULTS: When examined with a short TI inversion-recovery sequence with different TEs, the median nerve showed a 46-175% increase in signal intensity between 0 degrees and 55 degrees and an increase in mean T2 from 47.2 to 65.8 msec. When examined in 5 degrees to 10 degrees increments from 0 degrees to 90 degrees, the median nerve signal intensity changed in a manner consistent with the magic angle effect. No significant change was observed in skeletal muscle. Ulnar and sciatic nerves also showed changes in signal intensity depending on their orientation to B(0). Components of the brachial plexus were orientated at about 55 degrees to B(0) and showed a higher signal intensity than that of nerves in the upper arm that were nearly parallel to B(0). A reduction in the change in signal intensity in the median nerve with orientation was observed in the two patients with rheumatoid arthritis. CONCLUSION: Signal intensity of peripheral nerves changes with orientation to B(0). This is probably the result of the magic angle effect from the highly ordered, linearly orientated collagen within them. Differences in signal intensity with orientation may simulate disease and be a source of diagnostic confusion.

Magnetic resonance imaging of periosteum with ultrashort TE pulse sequences.

PURPOSE: To assess the values of pulse sequences with ultrashort echo times (0.08 msec) for detecting and characterizing periosteum. MATERIALS AND METHODS: Two normal volunteers aged 33 and 58 years and 12 patients aged seven to 55 years were studied. A total of 10 of the patients had contrast enhancement with intravenous Gadodiamide. Two ovine tibias were examined before and after the periosteum was stripped from the bone. RESULTS: High signal regions were observed adjacent to cortical bone in all parts of the skeleton imaged. They were generally more conspicuous after fat suppression and contrast administration. In the ovine tibia there was a reduction in the high signal normally seen at the surface of the bone after periosteal stripping. The detached periosteum produced a high signal. Mean T(2)* values for adult human periosteum ranged from 5.3 to 11.4 msec. After enhancement the signal intensity increased. In two patients with tibial fractures, increased periosteal signal was seen and this showed marked enhancement. Signals from periosteum could be simulated by fat, contrast-enhanced blood and artifacts. CONCLUSION: The periosteum can be visualized with ultrashort echo time pulse sequences in health and disease.

Magnetic resonance imaging of the liver with ultrashort TE (UTE) pulse sequences.

PURPOSE: To assess the feasibility of imaging the liver in volunteers and patients with ultrashort echo time (UTE) pulse sequences. MATERIALS AND METHODS: Seven normal controls as well as 12 patients with biopsy-proven generalized liver disease and three patients with focal disease were examined using pulse sequences with initial TEs of 0.08 msec followed by three later echoes, with or without frequency-based fat suppression. T(2)* values were calculated from regions of interest in the liver. RESULTS: Good image quality was obtained in each subject. There was a highly significant difference in the mean T(2)* values between the normal controls and patients with generalized liver disease (P = 0.001). T(2)* was significantly decreased in hemochromatosis (P = 0.002) and increased in cirrhosis (P = 0.04), compared with controls. T(2)* also correlated with functional status assessed by Child's grade (P = 0.001). A hepatocellular carcinoma showed reduced short T(2) components in the region of thermal ablation and evidence of a subcapsular hematoma which were not apparent with conventional imaging. CONCLUSIONS: Imaging of the liver with UTE sequences showed good image quality and tolerance of abdominal motion. T(2)* was specifically correlated with the presence of hemochromatosis, cirrhosis, and functional grade. Imaging of short T(2) relaxation components may provide useful information in disease.