Discriminating intrauterine growth-restricted piglets through near-infrared spectroscopy of amniotic fluids

Authors: FEUGANG, JEAN - Mississippi State University; SANTOS-RIVERA, MARINA - Mississippi State University; GRANT, ABIGAIL - Mississippi State University; PARK, SEONG - Mississippi State University; DEVOS-BURNETT, DERRIS - Mississippi State University; LEMLEY, CALEB - Mississippi State University; VANCE, CARRIE - Mississippi State University

Submitted to: Journal of Reproduction and Fertility
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/28/2019
Publication Date: 12/2/2019
Citation: Feugang, J., Santos-Rivera, M., Grant, A., Park, S., Devos-Burnett, D., Lemley, C., Vance, C. 2019. Discriminating intrauterine growth-restricted piglets through near-infrared spectroscopy of amniotic fluids. Journal of Reproduction and Fertility. 32(2):230.

Interpretive Summary: The occurrence of intrauterine growth restriction (IUGR) is a great concern in animal production systems, leading to substantial economic losses. The multifactorial causes of IUGR make their prediction before or during pregnancy a challenging task. The amniotic fluid, containing chemical secretions of both the developing fetus and the amnion-allantois membrane, may offer a unique opportunity to investigate the causes of IUGR in piglets. We were able to detect and discriminate chemical changes of the amniotic fluid from different fetuses using NIRS, indicating that this fluid contains molecules that may characterise IUGR in piglets. These observations will permit future investigations to identify specific molecules associated with IUGR that would lead to possible real-time corrective actions before birth.

Technical Abstract: Here we applied near-infrared spectroscopy (NIRS) as a rapid tool to discriminate the chemical composition of the amniotic fluid from IUGR fetuses at different gestational ages. The uteri of gravid sows were dissected at 60% (n'='6 sows) and 90% (n'='6 sows) of gestational age. Amniotic fluids were collected with a sterile 10-mL syringe connected to an 18-G needle under visual guidance, and samples without blood contamination were kept at 4°C while the fetus was measured. Samples were selected according to fetal weight and were classified as small (mean - 2 SD), medium (mean), and large (mean + 2 SD) sizes in each litter. Selected samples were kept at -80°C until NIRS transmittance spectra collection (350-2500'nm). At 60% of gestational age, male fetuses weighed 96'±'1.2, 133'±'1.7, and 159'±'1.9'g and female fetuses weighed 117'±'1.1, 137'±'1.8, and 152'±'1.8'g for small, medium, and large groups, respectively. The fetal weights at 90% of gestational age averaged 517'±'0.6, 757'±'0.6, and 912'±'0.7'g for males and 532'±'0.6, 718'±'0.9, and 911'±'0.6'g for females for small, medium, and large sizes, respectively. Chemometric analysis of the amniotic fluid spectra was applied to the wavelength range 1300-1600'nm using a balanced data set (n'='425) with an 80/20 calibration/validation distribution for linear discriminant analysis (LDA-PCA). Overall, we were able to discriminate samples according to fetal size and gestational age; LDA-PCA showed the highest size classification occurring with spectra derived from the small fetuses with an accuracy of 94.5-100% and 100% and 70.6% for the calibration and validation processes from 60% and 90% of gestational age, respectively. By contrast, when the analysis considered each sex at 60% of gestational age, the medium and small fetuses presented the highest size classification for females and males correspondingly. At 90% of gestational age, the large fetuses presented the best discrimination for both sexes.