Normal fetal lung volume measured with three-dimensional ultrasound

Authors


Abstract

Objectives

To construct reference intervals for fetal lung volumes measured longitudinally using three-dimensional (3D) ultrasound, and to evaluate the effect of gender on lung size.

Methods

This was a prospective, longitudinal study in the obstetric outpatient department of the VU University Medical Center, Amsterdam. Seventy-eight women with uncomplicated pregnancies were scanned three to four times at gestational ages of 18–34 weeks. 3D models of the lung were constructed using the ultrasound machine's software. After the infants were delivered the entire group was reanalyzed with regard to fetal gender. Centiles for the lung volumes of the entire group and for each gender separately were estimated using multilevel modeling.

Results

Charts and tables of right and left fetal lung volumes, using gestational age and estimated fetal weight as the independent variables, are presented. There was a significant difference in lung volume between male and female fetuses at each gestational age. Charts and tables of right and left fetal lung volumes for each gender at gestational ages of 18–34 weeks are also presented.

Conclusions

We present valid references for volumetric measurements of the right and left fetal lungs in male and female fetuses. The feasibility and reliability of fetal lung volume measurements using 3D ultrasound is good. Copyright © 2006 ISUOG. Published by John Wiley & Sons, Ltd.

Introduction

Fetal and neonatal pulmonary hypoplasia is defined as a reduction in the number of lung cells, airways and alveoli, resulting in a smaller and lighter organ. It may result from oligohydramnios, intrathoracic masses or skeletal and neuromuscular anomalies1. The incidence is 1.1 per 1000 live births and the associated mortality rate is about 70% (55–100%)1.

Conventional two-dimensional (2D) ultrasound, which is used routinely in obstetrics, has not proved sufficiently reliable to be used in the clinical diagnosis and management of pulmonary hypoplasia1–8. However, recent studies have indicated the value of three-dimensional (3D) ultrasound. This technology has the same advantages as 2D ultrasound of cost-effectiveness, ease and speed of use, and patient acceptability, but, in addition, enables visualization of perpendicular planes simultaneously (multiplanar imaging) and measurement of the volume of different organ systems (volume rendering)9–17.

Recent studies indicate that 3D ultrasound can be used in the estimation of fetal lung volume and that measurements are reproducible up to 30 weeks' gestation18–26. It is also anticipated that 3D ultrasound may be useful in detecting pulmonary hypoplasia prenatally23, 26–28. In order to determine the degree of pulmonary hypoplasia, nomograms of the right and left lung are required. In children a relationship between gender, age and growth in predicting lung volume and function has been well described29, 30. Recently it has also been suggested that gender-related calculation by ultrasound allows optimized prediction of fetal weight31–34.

Until now nomograms of fetal lung volumes have been created mostly using cross-sectional data and none of the studies differentiated between male and female fetuses18–22, 24–27. We present charts and tables of fetal lung volumes from 18 to 34 weeks' gestation measured longitudinally using 3D ultrasound and evaluate the effect of gender on lung size.

Methods

Sample

Between November 2002 and June 2004, we enrolled into the study 78 women who visited the obstetric outpatient department of the VU University Medical Center at 18–22 gestational weeks. All pregnancies had a known gestational age by last menstrual period, confirmed by sonographic measurement of the fetal crown–rump length (CRL) at 8–12 weeks or fetal biparietal diameter and head circumference at 12–20 weeks; gestational ages at recruitment were 18 (n = 16), 19 (n = 16), 20 (n = 17), 21 (n = 14) and 22 (n = 15) weeks. Exclusion criteria were maternal complications (e.g. premature delivery) or medication that was likely to affect growth of fetal lungs (corticosteroids), oligohydramnios (amniotic fluid index < 5th centile) and the presence of fetal malformations or abnormal growth (estimated fetal weight < 5th centile or > 95th centile and/or abdominal circumference < 5th centile or > 95th centile). Women with a pacemaker were also excluded because the transmitter necessary for the lung volume measurements could disturb the function of the pacemaker. The mean maternal age was 35 (range, 23–41) years. Seventy-two pregnancies had been conceived spontaneously, three resulted from in-vitro fertilization, two were conceived after intracytoplasmic sperm injection, and one was the result of intrauterine insemination.

No patients were lost to follow up. Infant birth weights were within normal ranges (> 10th and < 90th centile for gestational age), Apgar scores at 1 and 5 min were > 7 and none of the infants developed respiratory distress in the neonatal period. Fetal gender was not known before delivery so only after the children were delivered was the entire group analyzed again with regard to fetal gender.

The hospital ethics committee approved the study protocol and all women gave their informed consent before the first examination.

Measurements

Fetal biometry and lung volumes were measured three to four times during each pregnancy, with a mean interval of 4 (range, 3–5) weeks between measurements. All examinations were performed by the same examiner using a Technos MX (Esaote, Maastricht, The Netherlands) ultrasound machine equipped with a 3–5-MHz transabdominal annular array probe and with an integral 3D program. Images were acquired by moving the transducer over the maternal abdomen to visualize the entire fetal chest in a transverse plane from the neck to below the level of the diaphragm. 3D ultrasound was performed using a free-hand scanning technique with a position sensor attached to the transducer. Two to five volumes were acquired for each patient and stored on removable magneto-optical disks for offline analysis. Only when the diaphragm, clavicle and lung contours were visible were the volumes included in the final analysis. In order to standardize measurements of the lung, upper and lower anatomical limits were respectively set at the level of the fetal clavicles and at the dome of the diaphragm in the transverse and sagittal planes. The outline of each lung was traced manually on 5–15 slices in 5–10 min. From these 2D outlines the computer program automatically constructed and rendered a 3D model of the lung, and the computer's software calculated the volume of this 3D model in mL.

To determine the relationship between the lung volumes and the estimated fetal weight, we measured the biparietal diameter, head circumference, abdominal circumference and femur length, from which estimated fetal weight was derived using the Hadlock formula35.

To assess intraobserver variability the stored right and left lung volumes of each examination was remeasured on two further separate occasions by the same examiner at each gestational week (18–34 weeks). To assess interobserver variability, 24 scans, selected randomly, were measured three times by a second examiner.

Statistical analysis

Lung development over time (i.e. growth curves) was analyzed with multilevel modeling (or random coefficient analysis)36, 37. With this method, the individual regression curves were estimated and combined to obtain the population reference intervals for fetal lung volume. Multilevel analysis takes into account that successive measurements of each subject are related to each other, and by using this technique, it is not necessary to have equal time intervals between the successive measurements and the same number of measurements for each subject. This study can thus be described as a ‘mixed longitudinal study’, i.e. subjects were measured over different time periods. Population reference intervals were estimated over a broader age range (or weight range) compared with each of the individual curves. All multilevel analyses were performed with MLwiN (Multilevel models project, Institute of Education, London, UK). Intraobserver and interobserver intraclass correlation coefficients were calculated using SPSS for Windows, version 11.5 (SPSS Inc., Chicago, IL, USA). Bland–Altman plots and 95% limits of agreement between observers for lung volume measurements were calculated.

Results

A total of 308 fetal lung scans from 78 different patients were recorded. Seventy-four patients were scanned four times and four patients were scanned three times. Of the 616 volumes (right and left lungs), 22 volumes (3.6%) were excluded, because part of the lung contour could not be identified due to unfavorable fetal position or fetal breathing movements: 594 volumes (478/486 (98.3%) volumes between 18 and 30 weeks' gestation, and 116/130 (89.2%) volumes after 30 weeks' gestation) were included in the final analysis. Of these, 586 were included in the analysis of fetal lung volume vs. estimated fetal weight because a small number of fetuses had an estimated weight between 2400 g and 2750 g and these data were excluded. Of the 78 infants, 39 were male and 39 were female. In the male group 150 fetal lung scans were recorded and 148 were recorded in the female group.

The best fits for the right and left lung volumes with gestational age (weeks) as the independent variable were second-order polynomial regression equations, while those with estimated fetal weight (g) as the independent variable were simple linear regression equations (Table 1). According to these equations, the mean right lung volume ranged from 4.41 mL at 18 weeks to 44.99 mL at 34 weeks, and from 3.99 mL (200 g) to 43.05 mL (2300 g) (Tables 2 and 4). The mean left lung volume ranged from 3.10 mL (18 weeks) to 33.72 mL (34 weeks) and from 3.30 mL (200 g) to 33.19 mL (2300 g) (Tables 3 and 5). The relationships between the lung volumes and gestational age and estimated fetal weight are plotted in Figure 1.

Figure 1.

The 2.5th, 5th, 10th, 50th, 90th, 95th and 97.5th centiles for (a,c) right and (b,d) left lung volume according to (a,b) gestational age and (c,d) estimated fetal weight. equation image, 97.5th; equation image, 95th; equation image, 90th; equation image, 50th; equation image, 10th; equation image, 5th; equation image, 2.5th.

Table 1. Equations of best fit for lung volume calculated from the rendered three-dimensional models of the lung
 Gestational age (GA) (weeks) as independent variableEstimated fetal weight (EFW) (g) as independent variable
  1. The formula used for the volume calculation was as follows: volume = sum of all drawn contours of vol I, where vol i = distance i × (A + Sqrt (A × B) + B)/3, and where A = area of contour-1 and B = area of contour-2 and distance i is the distance between the slices where the contour was drawn.

Right lung volume
 Mean26.694 − 3.236 × GA + 0.111 × GA20.275 + 18.598 × EFW
 SD5.338 − 0.693 × GA + 0.0259 × GA20.741 + 0.004131 × EFW
Right lung volume: male fetus
 Mean11.852 − 2.054 × GA + 0.089 × GA20.88566 + 0.01826 × EFW
 SD−3.352 + 0.08357 × GA + 0.008625 × GA2−0.0161 + 0.004527 × EFW
Right lung volume: female fetus
 Mean45.254 − 4.721 × GA + 0.138 × GA2−0.00331 + 0.01859 × EFW
 SD22.735 − 2.208 × GA + 0.0578 × GA2−0.596 + 0.005616 × EFW
Left lung volume
 Mean16.381 − 2.142 × GA + 0.078 × GA20.475 + 14.233 × EFW
 SD−2.197 + 0.04019 × GA + 0.007244 × GA20.725 + 0.002843 × EFW
Left lung volume: male fetus
 Mean17.00122 − 2.13631 × GA + 0.07829 × GA20.61163 + 0.01433 × EFW
 SD2.006 − 0.266 × GA + 0.01198 × GA20.192 + 0.002868 × EFW
Left lung volume: female fetus
 Mean15.5995 − 2.13631 × GA + 0.07829 × GA20.59694 + 0.01379 × EFW
 SD−2.496 + 0.0166 × GA + 0.00905 × GA20.460 + 0.003315 × EFW
Table 2. Fetal right lung volume (mL) at 18–34 completed weeks of gestation, according to gestational age
Weeks of gestationnPercentileSD
2.5th5th10th50th90th95th97.5th
18151.551.912.324.416.506.917.271.275
19161.822.262.755.287.818.318.741.543
20172.192.723.316.379.4310.0310.551.865
21162.683.314.027.6911.3512.0712.702.235
22263.284.024.879.2313.5914.4315.182.656
23153.984.855.8610.9916.1217.1218.003.130
24194.775.796.9712.9718.9720.1421.173.661
25145.676.868.2215.1722.1223.4824.674.240
26276.688.059.6017.5925.5827.1428.504.870
27207.819.3611.1420.2429.3431.1232.675.551
28149.0310.7912.8023.1133.4235.4337.196.286
291810.3612.3414.6026.2037.8040.0642.047.071
302211.7914.0116.5429.5142.4845.0247.237.911
312313.3415.8018.6233.0547.4850.3052.768.801
321114.9917.7120.8336.8152.7955.9058.639.742
331416.7319.7423.1740.7858.3961.8364.8310.737
341018.6021.8925.6744.9964.3168.0871.3811.783
Total297
Table 3. Fetal left lung volume (mL) at 18–34 completed weeks of gestation, according to gestational age
Weeks of gestationnPercentileSD
2.5th5th10th50th90th95th97.5th
18151.151.391.673.104.534.815.050.872
19161.191.531.903.845.786.166.481.181
20171.371.792.274.747.217.698.111.505
21161.672.192.785.808.829.419.931.842
22262.102.713.417.0110.6111.3111.922.194
23152.653.364.198.3812.5713.3914.102.556
24193.324.145.089.9014.7215.6616.482.939
25144.105.046.1111.5817.0518.1219.053.337
26275.016.087.2813.4219.5620.7621.833.745
27206.077.248.5715.4122.2423.5824.754.168
28147.248.5310.0017.5625.1126.5927.884.607
29188.529.9411.5619.8628.1629.7831.205.061
30229.9411.4913.2622.3231.3833.1634.705.525
312311.4813.1615.0824.9434.8036.7238.406.010
321113.1414.9617.0427.7138.3840.4642.286.505
331414.9116.8819.1330.6442.1544.3946.367.020
341016.8318.9421.3533.7246.0948.5150.617.541
Total297
Table 4. Fetal right lung volume (mL) at 18–34 completed weeks of gestation, according to estimated fetal weight (EFW)
EFW (g)nPercentileSD
2.5th5th10th50th90th95th97.5th
200230.480.921.423.996.567.077.501.566
300271.421.972.605.859.109.7410.281.980
400282.353.023.787.7111.6312.4113.072.393
500173.284.074.979.5714.1715.0815.862.806
600204.225.126.1511.4316.7117.7418.643.219
700165.156.177.3313.2919.2520.4121.433.633
800136.097.228.5115.1521.7823.0824.214.046
900157.028.279.7017.0124.3225.7527.004.459
1000167.969.3210.8818.8726.8628.4229.784.872
1100118.8910.3712.0620.7329.4031.0932.575.286
1200109.8211.4213.2422.5931.9433.7635.365.699
1300910.7612.4714.4224.4534.4736.4338.146.112
14001211.6913.5215.6126.3137.0139.1040.936.525
15001312.6314.5716.7928.1739.5541.7743.716.939
1600813.5615.6217.9730.0342.0944.4446.507.352
1700814.5016.6719.1531.8944.6247.1149.287.765
18001415.4317.7220.3433.7547.1649.7852.078.179
1900616.3618.7721.5235.6149.7052.4554.868.592
2000717.3019.8222.7037.4752.2455.1257.649.005
2100518.2320.8723.8839.3354.7857.7960.439.418
2200619.1721.9225.0741.1957.3160.4663.219.832
2300420.1022.9726.2543.0559.8563.1366.0010.249
2400521.0424.0227.4344.9162.3965.8068.7810.658
Total293
Table 5. Fetal left lung volume (mL) at 18–34 completed weeks of gestation, according to estimated fetal weight (EFW)
EFW (g)nPercentileSD
2.5th5th10th50th90th95th97.5th
200230.410.771.183.305.425.846.191.291
300271.191.632.144.737.327.828.271.582
400281.982.503.106.159.209.8010.321.862
500172.773.374.067.5711.0811.7812.372.143
600203.554.235.009.0012.9913.7614.452.434
700164.345.105.9710.4214.8715.7416.502.714
800135.125.966.9211.8416.7617.7218.563.000
900155.916.837.8813.2718.6619.7020.633.286
1000166.707.708.8414.6920.5421.6822.683.566
1100117.488.569.7916.1122.4323.6624.743.852
1200108.279.4310.7517.5424.3325.6426.814.138
130099.0510.2911.7118.9626.2127.6328.874.423
1400129.8411.1612.6720.3828.0929.6130.924.704
15001310.6312.0313.6321.8129.9931.5932.994.990
1600811.4112.8914.5823.2331.8833.5735.055.275
1700812.2013.7615.5424.6533.7635.5537.105.556
18001412.9814.6216.4926.0835.6737.5339.185.847
1900613.7715.4917.4527.5037.5539.5141.236.128
2000714.5716.3618.4128.9239.4341.4943.276.408
2100515.3417.2219.3630.3541.3443.4745.366.699
2200616.1418.0920.3231.7743.2245.4547.406.980
2300416.9318.9621.2833.1945.1047.4349.457.260
2400517.7019.8122.2334.6146.9949.4151.527.551
Total293

Using gestational age as the independent variable, the entire group was then reanalyzed, calculating new equations for male and female fetuses separately (Table 1). According to these equations, the mean right lung volume ranged from 3.72 mL at 18 weeks to 44.90 mL at 34 weeks for the male fetus and from 4.99 mL at 18 weeks to 44.27 mL at 34 weeks for the female fetus (Tables 6 and 8). The mean left lung volume ranged from 3.91 mL at 18 weeks to 34.87 mL at 34 weeks for the male fetus and from 2.51 mL at 18 weeks to 33.47 mL at 34 weeks for the female fetus (Tables 7 and 9). The relationships between the lung volumes and gestational age are plotted in Figure 2.

Figure 2.

The 2.5th, 5th, 10th, 50th, 90th, 95th and 97.5th centiles for (a) right and (c) left lung volume of male fetuses and (b) right and (d) left lung volume of female fetuses. equation image, 97.5th; equation image, 95th; equation image, 90th; equation image, 50th; equation image, 10th; equation image, 5th; equation image, 2.5th.

Table 6. Right lung volume (mL) in male fetuses at 18–34 completed weeks of gestation, according to gestational age
Weeks of gestationnPercentileSD
2.5th5th10th50th90th95th97.5th
18101.621.882.183.725.255.555.820.936
1961.962.332.764.967.167.587.961.339
2082.432.923.486.379.269.8210.311.760
2193.043.664.367.9711.5812.2812.902.199
22153.794.535.389.7414.1014.9515.692.658
2354.675.556.5511.6916.8317.8318.713.133
24105.696.717.8713.8219.7720.9321.953.628
2576.878.029.3516.1322.9124.2325.394.133
26168.169.4710.9618.6126.2627.7529.054.663
2779.6211.0712.7421.2829.8131.4832.945.204
28811.2112.8214.6724.1233.5735.4137.035.763
29912.9414.7116.7427.1437.5439.5641.346.339
301314.8116.7518.9730.3341.6943.9145.856.929
31816.8418.9421.3633.7146.0648.4750.587.533
32719.0021.2823.8937.2650.6353.2455.528.153
33621.3123.7726.5840.9955.4058.2160.678.786
34623.7726.4129.4344.9060.3763.3966.039.434
Total150
Table 7. Left lung volume (mL) in male fetuses at 18–34 completed weeks of gestation, according to gestational age
Weeks of gestationnPercentileSD
2.5th5th10th50th90th95th97.5th
18101.451.762.113.915.716.066.371.097
1961.802.162.574.676.777.197.541.283
2082.252.673.155.598.038.518.931.490
2192.823.313.856.669.4710.0210.491.712
22153.514.064.687.8911.1011.7312.271.957
2354.304.925.639.2812.9313.6414.262.224
24105.195.906.7010.8214.9415.7516.452.513
2576.196.987.8812.5217.1518.0618.852.826
26167.298.189.1914.3819.5720.5821.463.163
2778.499.4810.6016.3922.1823.3124.293.528
2889.7910.8912.1418.5624.9826.2427.333.916
29911.1912.4013.7820.8927.9929.3830.594.332
301312.6714.0115.5423.3731.2032.7334.074.775
31814.2515.7217.4026.0134.6236.3037.775.250
32715.9217.5319.3728.8138.2540.0941.705.755
33617.6719.4321.4431.7642.0844.0945.856.291
34619.5121.4323.6234.8746.1148.3150.236.857
Total150
Table 8. Right lung volume (mL) in female fetuses at 18–34 completed weeks of gestation, according to gestational age
Weeks of gestationnPercentileSD
2.5th5th10th50th90th95th97.5th
1861.111.602.154.997.838.388.861.730
19101.642.112.645.378.108.639.091.663
2092.192.683.226.038.849.399.861.712
2172.783.303.906.9710.0410.6411.161.872
22113.363.974.658.1811.7112.4013.002.150
23103.974.685.499.6713.8414.6615.372.546
2494.595.456.4311.4416.4517.4318.283.056
2575.236.267.4413.4819.5220.6921.723.681
26115.907.138.5515.8023.0524.4625.704.421
27136.568.049.7318.3927.0528.7430.225.281
2867.269.0111.0121.2631.5133.5135.266.250
2997.9610.0212.3624.4036.4438.7940.847.339
30108.6811.0813.8027.8241.8344.5746.968.546
31149.4312.1915.0831.5247.4250.8553.619.862
32410.1813.3516.9635.4954.0257.6460.8011.298
33810.9614.5618.6739.7460.8164.9368.5212.849
34411.7615.8220.4644.2768.0872.7276.7814.515
Total148
Table 9. Left lung volume (mL) in female fetuses at 18–34 completed weeks of gestation, according to gestational age
Weeks of gestationnPercentileSD
2.5th5th10th50th90th95th97.5th
1860.861.071.302.513.713.954.160.735
19100.841.141.493.275.055.405.701.087
2090.931.341.804.196.577.047.451.454
2171.121.652.235.268.298.889.401.847
22111.452.082.806.4910.1810.9011.532.250
23101.892.643.507.8812.2613.1213.872.673
2492.443.324.319.4214.5315.5316.403.115
2573.114.125.2611.1216.9818.1319.133.574
26113.905.036.3312.9819.6320.9222.064.053
27134.806.087.5314.9922.4523.9125.184.548
2865.827.248.8517.1625.4627.0928.505.064
2996.958.5210.3119.4928.6730.4632.035.597
30108.209.9211.8921.9732.0534.0235.746.148
31149.5711.4513.6024.6135.6237.7739.656.714
32411.0513.0915.4327.4139.3941.7243.777.304
33812.6514.8617.3930.3643.3345.8648.077.908
34414.3616.7519.4833.4747.4650.1952.588.531
Total148

The entire group was also reanalyzed, producing new equations for male and female fetuses separately, using the estimated fetal weight as the independent variable (Table 1). The mean lung volume for the male fetus ranged from 4.54 mL (200 g) to 42.88 mL (2300 g) for the right lung and from 3.48 mL (200 g) to 33.57 mL (2300 g) for the left lung. For the female fetus mean lung volume ranged from 3.71 mL (200 g) to 42.75 mL (2300 g) for the right lung and 3.35 mL (200 g) to 32.31 mL (2300 g) for the left lung.

Comparing the mean right and left lung volumes of the male and female fetuses with the mean right and left lung volumes of the total group, the differences for gestational age and estimated fetal weight became apparent (Figure 3). The mean difference between the male and female lung volume for gestational age was 4.3% and for estimated fetal weight it was only 1.4%. In both male and female groups the Wald test indicated a statistical correlation of both the right and the left lung volumes with each gestational age, but not with estimated fetal weight.

Figure 3.

Differences between the mean lung volume of the total group and the mean gender-specific lung volume for (a) gestational age and (b) estimated fetal weight. equation image, male fetus, left lung; equation image, male fetus, right lung; equation image, female fetus, left lung; equation image, female fetus, right lung.

Combining the centiles of the male and female lung volumes in one graph, allowed the differences to be well demonstrated (Figure 4). The upper limit of the right female lung volume (97.5th centile) was lower than that of the male lung volume until 29 weeks' gestation; after 29 weeks the female lung exceeded the male lung. The 50th centile of the right female lung volume was somewhat lower than that of the male lung volume, but both volumes were equal at around 34 weeks' gestation. The lower limit (2.5th centile) of the right lung volume was the same for female and male fetuses until 21–23 weeks, when the lower limit of the male right lung volume showed a steeper incline compared with the female right lung volume. The upper limit of the female left lung volume was lower than the male one until 25 weeks, but was higher from 25 weeks onwards. Both the 5th and 50th centiles of the left female lung were lower than those of the male lung from 18 to 34 weeks' gestation.

Figure 4.

Gender-specific nomograms (2.5th, 50th and 97.5th centiles) for (a) right and (b) left lung volumes. equation image, 97.5th♀; equation image, 97.5th♂; equation image, 50th♂; equation image, 50th♀; equation image, 2.5th♂; equation image, 2.5th♀.

To investigate whether or not the variability differed for males and females after 30 weeks' gestation, we added an interaction between a dichotomous time-indicator (< 30; ≥ 30) and gender to the regression analysis. The P-values of these interactions were an indication of whether the variability after 30 weeks was different for males and females. The right lung had a P-value of 0.027, while the left lung had a P-value of 0.042. Both interaction terms were positive, so there was indeed a significant difference in variability after 30 weeks between males and females; females had greater variability than did males.

The intraclass correlation coefficients between three repeated measurements performed by Observer 1 (intraobserver correlation) at each gestational week (18–34 weeks) were 0.97 (range, 0.93–0.99) for the right and 0.95 (range, 0.90–0.98) for the left lung volumes, using 891 measurements for each lung. Observer 2 had a value of 0.99 for both the right and left lungs, using 72 measurements for each lung. The correlation coefficients between measurements made by different observers (interobserver correlation) were 0.97 for the right and 0.96 for the left lung, using 72 measurements for each lung. The 95% limits of agreement between the two observers were −1.55 to 3.81 mL for the left lung and −3.5 to 3.95 mL for the right lung (Figure 5).

Figure 5.

Difference in right lung volume measurements between Observer 1 (obs 1) and Observer 2 (obs 2) plotted against their average, (Observer 1 + Observer 2)/2.

Discussion

Conventional 2D ultrasound has been used to evaluate fetal breathing movements, thoracic circumference, lung length and the ratio of thoracic circumference to abdominal circumference in order to predict lung volume. However, these parameters were not sufficiently sensitive or specific for clinical management in cases of pulmonary hypoplasia1–8. It has been shown that measuring lung volume with magnetic resonance imaging (MRI) has benefits over conventional 2D ultrasound38–45, particularly with the new high-speed systems, which have clearer MR images with fewer motion artifacts. Limitations of MRI in daily practice are the high costs and limited acceptance by pregnant women. Recent studies report that it is also possible to measure fetal lung volumes reliably with 3D ultrasound18–26; advantages include the cost-effectiveness, ease and speed of use, and patient acceptability.

After 30 weeks' gestation it is more difficult to measure lung volumes because of poor image quality. An unfavorable position of the fetus, increased ossification of the spine and ribs, and motion artifacts due to fetal breathing movements can make visualization of the lungs, diaphragm or clavicles more difficult. Most studies had success rates between 85% and 90%19, 20, 24–26. In their study of lung volume measurements between 15 and 40 weeks' gestation, Osada et al.26 found that 87.2% were eligible for analysis. Before 20 weeks and after 34 weeks of gestation, 25% of the volumes were excluded because of poor image quality due to prone presentation of the fetal spine, while between 20 and 34 weeks only 2.9% had to be excluded. Sabogal et al.25 studied the fetal lungs between 20 and 30 weeks of gestation and excluded 8.2% of the volumes because of poor image quality.

In this study we had to exclude only 3.6% of the volumes. Before 22 weeks' gestation we had no difficulties in acquiring good image quality. Most difficulties occurred after 30 weeks' gestation; 14 of the 130 volumes (10.8%) were excluded. Poor image quality was due mostly to an unfavorable position of the fetus and with advancing gestation increased ossification became a greater problem. Although we repeated the scanning process if fetal breathing movements were present, a few volumes still had to be excluded because of inadequate visualization of the fetal thorax. Overall, we were able to acquire good image quality in most (96.4%) cases.

It was not only the feasibility of the fetal lung volume measurements, but also the reliability, that was good. The intraobserver variability at each gestational week (18–34 weeks) was < 3% and 5% for the right and left lung volumes, respectively. The 95% limits of agreement between the two observers were small for both right and left lungs.

In this study, we measured the lung volumes in the transverse plane, of which the accuracy has been confirmed in a study by Pohls and Rempen19. The lung volumes were measured directly between the clavicles and the diaphragm; outside these limits it is difficult to visualize lung tissue clearly. This approach is similar to that used by other investigators19, 21–26. Some studies measured lung volumes indirectly, by subtracting the heart volume from the thoracic volume18–20. The disadvantages of this indirect method are that not only lungs but also blood vessels and thymus gland are included in the thoracic volume measurements and that only the total lung volume can be determined rather than the right and left lungs separately. Direct and indirect methods were evaluated by Pohls and Rempen19 and although there was a high correlation, there was also a significant difference between the two approaches.

We compared our volume graphs with graphs from other studies which also used the direct measuring method21. The longitudinal study by Bahmaie et al.21 reported 3.22 mL at 18–20 weeks to 24.77 mL at 33–34 weeks of gestation for the right lung and 2.54 mL at 18–20 weeks to 21.06 mL at 33–34 weeks of gestation for the left lung, using data from 235 scans. Our values were larger at all gestational ages, with values of 6.37 mL at 20 weeks to 44.99 mL at 34 weeks of gestation for the right lung and 4.74 mL at 20 weeks to 33.72 mL at 34 weeks of gestation for the left lung. However, our graphs were similar to those of Rypens et al.41, who used fast-spin echo MRI.

We also compared our graphs with fetopathological standards given by Langston et al.46 and more recently by De Paepe et al.47, although we realize that it is difficult to compare postmortem pathological studies with a sonographic study. The volumes measured in our study were still below the volumes obtained from the pathological studies, but correlated better than did earlier 3D ultrasound studies19, 21–26, 34.

A constant linear relation was observed between left and right fetal lung volumes throughout pregnancy. The left and right lung volumes corresponded, respectively, to 44% and 56% of the total lung volume. These results are also in agreement with postmortem data obtained in infants and fetuses47, 48. We also showed an approximately seven-fold increase in right and left fetal lung volume during the second half of pregnancy (20–34 weeks). This increased rate is consistent with findings in other 3D ultrasound studies19, 21, 26.

It is well known that female fetuses, on average, weigh less than do male fetuses at any gestational age; the mean weight of male fetuses exceeds that of female fetuses by 2–3%. The gender difference in fetal growth appears to be rather pronounced before the third trimester and relatively less marked towards term. It has therefore been suggested that the use of gender-specific nomograms may improve prenatal assessment of fetal growth30–33. In this study a significant difference in the left and right lung volumes between male and female fetuses for each gestational age was demonstrated according to the Wald test. The mean lung volume of male fetuses was 4.3% larger compared with that of female fetuses. The difference was not significant when the lung volumes were plotted against the estimated fetal weight. Like fetal weight the differences in fetal lung volumes are more apparent in the second trimester and less marked in the third trimester. Langston et al.46 noted no differences in lung volumes between male and female fetuses in a histopathological study. The results were expressed as correlation coefficients with CRL obtained postmortem. The CRL was then translated to the corresponding gestational age in weeks. Because the difference in growth between male and female fetuses was not taken into account, it is not surprising that there was no difference between male and female fetal lung volume. Although we did observe a significant difference in the left and right lung volumes between male and female fetuses, it needs to be investigated whether these small differences are clinically relevant.

Most studies18–20, 22–26, 34, except for that by Bahmaie et al.21, used cross-sectional data to create reference ranges of the fetal lungs. The means of these reference ranges indicate the average size of the fetal lung at a certain gestational age. However, to create valid references concerning growth it is not possible to use these single measurements of volume. Our charts were constructed using measurements on three or four occasions (longitudinal) and are therefore valid references with respect to the growth of the fetal lung. It should be borne in mind that the outer centiles are less accurate compared with the 50th centile because the number of these measurements in the total group and especially in the male and female groups was small. The potential of 3D ultrasound and these new references in predicting pulmonary hypoplasia needs to be evaluated.

Ancillary