Course Director

September 28-30, 2018

Wesley Lee, M.D. Professor, Department of Obstetrics and Gynecology Division Director, Women’s and Fetal Imaging Baylor OBGYN - Texas Children’s Hospital Houston, Texas

wesley.lee@bcm.edu

Why is Fetal Soft Tissue Assessment Important?

Chicago OB/GYN Ultrasound SymposiumUniversity of ChicagoHotel Chicago - September 29, 2018

1

Financial Disclosures

I have no relationships with commercial companiesthat could be perceived as a conflict of interest although the following relationships are disclosed:

Samsung Limited Research Support

Toshiba Technology Consultant

Philips Ultrasound: Technology Consultant

2

Learning Objectives

After completing this presentation, the participant should be able to:

1. Describe the rationale for evaluating fetal soft tissue

2. Discuss conventional approaches for soft tissue assessment using 2DUS

3. Explain how 3DUS can be used to measure fractional limb volume

4. Summarize advantages and limitations for using FLV to improve the precision of estimated fetal weight

3

Why is Fetal Soft Tissue Important?

• soft tissue assessment improves precision of EFW and now addsanother key nutritional component to the weight estimation process

• fetal growth is a complex process and should be characterizedusing a combination of skeletal and soft tissue parameters

4

159 Healthy Term and Preterm Newborns87 boys and 72 girls

Dual-energy X-ray Absorptiometry (DXA)

5

Term Infants: 42.9% of newborns < 10%body fat were classified as AGA

9.9% AGA newborns had < 10% body fat

6

• Fetal thigh circumferenceDeter RL, et al (1983)

• Subcutaneous tissue - arm and leg Jeanty P, et al. (1985)

• Fetal thigh and calf circumferencesVintzileos A, et al. (1985)

• Fetal limb fat and lean body massBernstein IM, et al (1997)

• Fetal abdomen subcutaneous tissueHill LM, et al (1992)Petrikovsky BM, et al (1997)Gardeil F, et al (1999)

• Cheek-cheek diameterAbramowicz JS, et al, (1991 & 1993)

• Fetal buttocksMatsumoto M, et al (2000)

• Fetal thigh soft tissue thicknessScioscia M, et al (2008)

Fetal Soft Tissue Parametersusing Ultrasonography

7

“In animal studies, regardless of the model used, FGR fetuses are unable to

deposit normal fat stores in late gestation. When FGR babies are

examined after delivery, they show all the clinical signs of reduced fat stores.”

Sharma D et al. Clin Med Insights Pediatr. 2016;10:67-83,

Battaglia, FC. NeoReviews 2003; 4: e91

“Thus, if fat concentration could be estimated fairly precisely during fetal life with ultrasonographic estimates of fat-to-lean ratios in various sites, this

might help distinguish the “normal small” from the FGR fetuses.

8

Neonatal Nutrition Score - Dr. Reba Hill9

Obstet Gynecol 2001;185:839-844

10

Fetuses with intrauterine growth restriction have reduced subcutaneous fat and lean body mass compared to normal controls.

Subcutaneous fat concentrations decreases are greater when compared to decreases in lean body mass

11

28.0 weeks

34.5 weeks

12

Neonatal Thighs

Growth Restriction - 2845 grams39.7 weeks, menstrual age

Macrosomia - 4368 grams38.4 weeks, menstrual age

13

MTFM

MTLM

Red Dot Marks Cross Sectional View of Femur Diaphysis

MTFM = Mid-Thigh Fat Mass (Subcutaneous Fat; cm )

MTLM = Mid-Thigh Lean Mass (Muscle and Bone; cm )

Mid-Thigh Area (i.e. Fat area + Lean Mass Area; cm )MTA =

MTFM = MTA - MTLMMTLM = MTA - MTFM

Calculations

ThC = Mid-Thigh Circumference (cm)

Limb Tracing for ThCand Mid-Thigh Area (MTA)

2

2

2

Fetal Thigh 2D Soft Tissue Parameters - NICHDMid-Point of Diaphysis Defined Using 3D Multi-Planar Imaging

MTSCTT = Maximum Thigh Subcutaneous Tissue Thickness; cm)

MTSCTT

14

Fractional Limb Volume

Limb Subvolume based on 50% of long bone diaphysis length

Fractional Arm Volume Fractional Thigh Volume

Lee W, et al. Fractional limb volume: a soft tissue parameter of fetal body composition: validation, technical considerations, and normal ranges during pregnancy. Ultrasound Obstet Gynecol 2009;33:427-440

15

Initial 3D Acquisition - Fetal Thigh16

17

Fractional Thigh Volume - 20 weeks

18

Fractional Limb Volume - Technical Considerations

Edges of the LongBones Have Shadowed Soft Tissue Borders

19

Fractional Limb Volume - Technical Considerations

Use Color Filter for Soft Tissue Contrast

Use Adequate Acquisition Sweep Angle

20

Avoid Motion Artifacts!

21

Fractional Limb VolumeFetal Soft Tissue Development

22

23

Fractional Limb Volume - Reproduciblity Study

Lee W, et al. Ultrasound Obstet Gynecol 2009;33: 427-440

2.0 ± 4.2% [95% LOA -6.3 to 10.3%]2.2 ± 4.2% [95% LOA -6.0 to 10.5%]

1.9 ± 4.9% [95% LOA -11.6 to 7.8%] -2.0 ± 5.4% [95% LOA -12.5 to 8.6%]

24

Fractional Arm Volume - Normal Ranges

Lee W, et al. Ultrasound Obstet Gynecol 2009;33: 427-440

25

Fractional Thigh Volume - Normal Ranges

Lee W, et al. Ultrasound Obstet Gynecol 2009;33: 427-440

26

27

28

Fractional Limb VolumeFetal Weight Estimation

29

Obstet Gynecol 1997; 90: 331-9

30

R2 = 0.46

31

EFW - Technical Limitations

Most prediction models do not include soft tissue parameters for fetal weight estimation . . .

• soft tissue parameters difficult to standardize • 3DUS fetal limb measurements can be time consuming • soft tissue borders poorly visualized at ends of long bones

From Barker DJP. Mothers, Babies, and Health in Later Life2nd Edition. Churchill Livingston, 1998

32

�33

How do fetal size parameters and actual birthweight correlate with newborn infant body compostion?

33

Air Displacement Plethysmography

Density = Mass

Volume

Two Compartment ModelFat

Fat-Free Mass

34

n = 87 fetus-newborn pairs

Lee W, et al. Fetal growth parameters and birth weight: their relationship to newborn infant body composition. Ultrasound Obstet Gynecol 2009;33:441-446

35

• 47 mother-infant pairs studied • pregnant women scanned (33-38 weeks)• infant body composition study (24 - 72 hrs)

J Nutri 2009;139:1772-8

36

**

*

*

*

*

*

*

*

Relationships Between

Fetal and Infant Biometry

to Newborn Adiposity

Moyer-Mileur LJJ Nutri 2009;139:1772-8

37

“We conclude that fetal growth biometrics determined by 2D US do not provide a

reliable assessment of % BF in term infants.”

Moyer-Mileur LJ, et al.J Nutri 2009;139:1772-8

38

BW Predictions - Accuracy versus Precision

Predicted BW - Actual BWActual BW

x 100

Standard Deviation of Percent Differences

• •••••••

• •••

precise

precise not accurate

not precise not accurateaccurate

39

Best EFW precision (lowestrandom errors) occurred with

volume-based weightmodels

2013

* systematic error significantly different from zero, one sample t-test

Fetal Weight Estimation Model

Birth WeightAll Infants (g)

Birth Weight < 2000 g

Birth Weight 2000-4000 g

Birth Weight > 4000 g

BPD, AC, FDLOriginal Hadlock (OH2)

4.9 ± 8.8*(n = 158)

4.9 ± 10.6*(n = 28)

4.4 ± 8.2*(n = 100)

6.7 ± 8.6*(n = 30)

BPD, AC, FDLModified Hadlock (MH2)

1.1 ± 8.4(n = 158)

1.0 ± 10.0(n = 28)

1.2 ± 8.0(n = 100)

0.5 ± 8.3(n = 30)

BPD, AC, TVolNew Model 6

1.9 ± 6.6*(n = 156)

0.4 ± 7.8(n = 28)

1.5 ± 6.4*(n = 98)

4.3 ± 5.8*(n = 30)

40

“Sonographically based fetal limb volumes, especially fractional thigh volume, reflect neonatal fat mass and are better correlated with birth weight.”

J Ultrasound Med 2009; 28:309-315

41

Model Derivation Group (n = 100) BW = 774.744 + 32.568 x TVol

Model Validation Group (n = 50)

Mean % Difference Hadlock -3.1 ± 7.8% New Model 0.2 ± 5.5%

J Med Assoc Thai 2009;92: 1580-5

42

Prospective Study (n = 190)

Mean % Difference EFW ± 5% EFW ± 10%

Modified Hadlock (BPD, HC, AC, FL)

-2.5 ± 6.9% 46.8 83.7

Volume Based Model (BPD, AC, TVol)

-0.4 ± 4.8% 74.2 95.8

43

“strong correlation between 3D thigh volumemeasurements at 33 weeks gestation and birth weight”

“adding TVol measurements to EFW at 38 weeks at38 weeks improves accuracy and reduces error rate”

44

Longitudinal prospective cohort study of 115 unselected pregnancies

“Fractional thigh volume measurements offer some improvement over 2-dimensional biometry for the detection

of late-onset fetal growth restriction at 34-36 weeks.”

Am J Obstet Gynecol. 2017;217:453.e1-453.e12.

45

5D Limb Vol Tool

Semi-Automated Fractional Limb Volume Measurements

46

Automated Fractional Limb Volume Demo47

Semi-Automated FLV With Manual Editing48

Menstrual Age (n = 50)

Prospective Study - Samsung WS80 with EliteComputer Assisted Fractional Limb Volume

Reproducibility

Time Efficiency

Mack LM, et al. J Ultrasound Med 2016; 35:1573-8

49

Semi-Automated Fractional Limb VolumeBland-Altman Analysis

1.7 ± 4.6% (95% LOA: -7.3 to +10.7%) -0.0 ± 3.8% (95% LOA: -7.5 to +7.5%)

Mack LM, et al. J Ultrasound Med 2016; 35:1573-8

50

Time Required to Complete FLV MeasurementsManual vs Automated Fractional Limb Volume (n = 10)

Seco

nds

0

50

100

150

200

Fractional Arm Volume Fractional Thigh Volume

Manual Semi-Automated

p < 0.001 p = 0.005

2 min36 seconds

31 seconds 38 seconds

3 min4 seconds

Mack LM, et al. J Ultrasound Med 2016; 35:1573-8

51

•50third-trimesterfetusesscannedwithin4daysofdelivery• Hadlock(BPD,AC,FDL)comparedtoLee(BPD,AC,TVol)BWpredicIons• 5DLimbVoltechnology-SamsungWS80

J Ultrasound Med 2017; 36: 1649-1655

52

Mostaccurateweightpredic1onsresults-HadlockModel

Mostpreciseweightpredic1onsresults-LeeModel

AutomatedFrac1onalLimbVolume

MeanBW=3335gramsat39.4±1.3weeksgestaIon

53

Correc1onFactorsforSystema1cErrorsAutomatedFrac1onalLimbVolume

AVol=1-(-0.0907)=1.0907 TVol=1-(-0.0523)=1.0523)

Late3rdTMFetuses

54

AutomatedFetalWeightEs1ma1on:AMul1centerValida1onUsingFrac1onalLimbVolume

Prospective study of 600 pregnancies

PrimaryObjec-ve-ComparetheaccuracyandprecisionofBWpredicIonsbasedon5DLimbVoltechnologyinpregnantwomenascomparedto2DBWpredicIonmodels.

Secondary Objec-ve - Characterize the performance of BW predicIons based on 5DLimbVoltechnologyinpregnantwomenwithsuspectedgrowthabnormaliIes.

a.BaylorCollegeofMedicineatTexasChildren’sHospitalPavilionforWomen(Houston,USA)b.MageeWomensHospital,Pil[sburghPAc.BeijingObstetricsandGynecologyHospital,(Beijing,China)d.ErlangenUniversityHospital(Hanover,Germany).

55

Fractional Limb Soft Tissue - Summary

Fetal soft tissue assessment requires development of ultrasound parameters that are reproducible and for which reference ranges have been established.

Fractional limb volume, when used with 2D biometry, improves the precision of EFW and now adds a soft tissue component to the weight estimation procedure.

The clinical value of fractional limb volume - either as a directly measured soft tissue parameter or for fetal weight estimation requires further investigation.

56

!57

57