Localized fetomaternal hyperglycemia: spatial and kinetic definition by positron emission tomography

Abstract

Background: Complex but common maternal diseases such as diabetes and obesity contribute to adverse fetal outcomes. Understanding of the mechanisms involved is hampered by difficulty in isolating individual elements of complex maternal states in vivo. We approached this problem in the context of maternal diabetes and sought an approach to expose the developing fetus in vivo to isolated hyperglycemia in the pregnant rat.

Methodology and principal findings: We hypothesized that glucose infused into the arterial supply of one uterine horn would more highly expose fetuses in the ipsilateral versus contralateral uterine horn. To test this, the glucose tracer [18F]fluorodeoxyglucose (FDG) was infused via the left uterine artery. Regional glucose uptake into maternal tissues and fetuses was quantified using positron emission tomography (PET). Upon infusion, FDG accumulation began in the left-sided placentae, subsequently spreading to the fetuses. Over two hours after completion of the infusion, FDG accumulation was significantly greater in left compared to right uterine horn fetuses, favoring the left by 1.9+/-0.1 and 2.8+/-0.3 fold under fasted and hyperinsulinemic conditions (p<10(-11) n=32-35 and p<10(-12) n=27-45) respectively. By contrast, centrally administered [3H]-2-deoxyglucose accumulated equally between the fetuses of the two uterine horns. Induction of significant hyperglycemia (10(3) mg/dL) localized to the left uterine artery was sustained for at least 48 hours while maternal euglycemia was maintained.

Conclusions and significance: This approach exposes selected fetuses to localized hyperglycemia in vivo, minimizing exposure of the mother and thus secondary effects. Additionally, a set of less exposed internal control fetuses are maintained for comparison, allowing direct study of the in vivo fetal effects of isolated hyperglycemia. Broadly, this approach can be extended to study a variety of maternal-sided perturbations suspected to directly affect fetal health.

Figure 1. A model for selective fetal exposure.

(a) The vast majority of blood flow to each uterine horn is supplied by each horn's uterine artery , which derives from the iliac artery. Therefore, infusate from a catheter appropriately placed in the left iliac artery should experience first pass exposure to fetuses in the left, but not right, uterine horn. The fetuses in the right uterine horn thus serve as internal controls. (b) In surgical approach “A”, ligatures (green) were tied around the left internal iliac artery just inferior to the uterine artery take off and around the femoral artery. Surgical approach “B” included additional ties (blue) around the left superior gluteal and hypogastric trunk arteries. (c) Experimental time-line. (d) Dynamic PET of the abdomen was initiated upon infusion of FDG, followed by a whole-body scan (WB) at 90-120 minutes, transmission and emission contamination scans (TE) at 120-160 minutes, and tissue collection after euthanasia at 160 minutes.

Figure 2. Final accumulation of glucose tracers.

(a) Per gram accumulation of ¹⁸F in various tissues at 160 minutes after FDG infusion. (* p<10⁻⁵ for left versus right, n = 76–79). Per fetus (b) and per placenta (c) accumulation of ¹⁸F, based on uterine horn position. The fetus/placenta closed to the cervix is counted as position 1. * p<0.05 left versus right. (d) Clearance of ³H into various tissues as measured after completion of the scan (n.s. not significant for left versus right). Per fetus (e) and per placenta (f) accumulation of ³H, based on uterine horn position.

Figure 3. Dynamic PET imaging.

(a) Representative dynamic PET images, shown as maximum intensity projections. Dynamic scans were centered on the abdomen. Minutes (“m”) after initiation of infusion are shown on each image. (b) SUV evolution for several anatomic structures. (c) SUV evolution in the left and right uterine horns.

Figure 4. PET whole-body imaging.

(a) Representative maximum intensity projection whole-body image. Color-intensity gradations are indicated from least to greatest. (b) Correlation between standardized uptake values (SUV) for volumes of interest as measured from the final whole body image (SUV_scan) and for the corresponding tissues as determined by gamma well counter measurement, dose administered, and weight of the rat and the tissue (SUV_well) (r = 0.82, n = 66, p<10^-16). Abbreviations: RU – right uterine horn; LU – left uterine horn.

Figure 5. Impact of early FDG pattern on differential glucose exposure.

(a) Impact of surgical approach on differential ¹⁸F (black bars) but not ³H (white bars) accumulation (* p<0.05 for difference between approaches, n = 5–6). (b) The ratio of ¹⁸F cranial versus caudal to the catheter tip is plotted over time for scans that exhibited a final left/right uterine horn ¹⁸F ratio of >1.6 (SEL) or <1.6 (NON). (*p<0.01, #p<0.001, n = 5–6).

Figure 6. Impact of infusate flow on differential glucose exposure.

(a) Accumulation of fluorescent microspheres (FMS) in placenta by laterality and position (1 is closest to cervix; z is closest to ovary) and tissues (L, liver; H, heart; B, brain). The number of FMS normalized to weight and total number, such that a tissue receiving an average number per gram would have a relative drainage of 1. (b) Relative accumulation of FMS between left and right, dividing the average number per left placenta by the average number per right placenta, among six individual scans. (c) Accumulation of FMS in placenta and tissues only among scans showing left-dominant FMS accumulation. (d) The ratio of FDG concentration cranial versus caudal to the catheter tip for scans that exhibited a left- or right-dominant FMS pattern (*p<0.05, n = 2–4). (e-f) SUV in left uterine (LU) versus right uterine (RU) horns for (e) left-dominant or (f) right-dominant FMS patterns. (g) Final accumulation of FDG in left and right fetuses and placenta in scans showing left- or right- dominant FMS accumulation. (*p<10^-4, for left versus right ¹⁸F, n = 19–25; ns p = not significant, n = 8–20).

Figure 7. Chronic glucose infusion.

Glucose was infused on gestational days #18-20 at 4 mg/min via the left artery catheter. Serum was sampled downstream of the infusion catheter tip (LU, black squares) and from the maternal tail tip (mat, open circles). *p<0.05.